USF Libraries
USF Digital Collections

Analysis of automatic vehicle identification technology and its potential application on the Florida turnpike


Material Information

Analysis of automatic vehicle identification technology and its potential application on the Florida turnpike
Physical Description:
iv, 143 leaves : ill. ; 28 cm.
Florida -- Dept. of Transportation
University of South Florida -- Center for Urban Transportation Research
USF Faculty and University Publications
University of South Florida, College of Engineering, Center for Urban Transportation Research
Place of Publication:
Tampa, Fla
Publication Date:


Subjects / Keywords:
Toll roads -- Automation -- Florida   ( lcsh )
Motor vehicles -- Automatic location systems   ( lcsh )
Florida Turnpike (Fla.)   ( lcsh )
non-fiction   ( marcgt )


Additional Physical Form:
Also issued online.
Statement of Responsibility:
by the Center for Urban Transportation Research.
General Note:
Prepared for the Florida Dept. of Transportation.
General Note:
"October 1990."

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 023228643
oclc - 25778759
usfldc doi - C01-00081
usfldc handle - c1.81
System ID:

This item is only available as the following downloads:

Full Text


Analysis Of Automatic Vehicle Identification Technology And Its Potential Application On The Florida Turnpike Technical Memorandum 2 Prepared for the Florida Department of Transportation by the Center for Urban Transportation Research October 1990


TABLE OF CONTENTS Sec tion Page T ABI..E OF CON"l"ENTS . . . . . . . . . . . . . . . . . . 1 UST OF T.ABLES . . . . . . . . . . . . . . . . . . . . . iii UST OF FIGURES . . . . . . . . . . . . . . . . . . . . iv I. II. Ill. IV. v. VI. INTRODUCTION AND OVERVIEW .................. SYNOPSIS OF TECHNICAL MEMORANDUM I ... ...... DESCRIPTION OF FLORIDA TURNPIKE SYSTEM ...... . A Existing System .................................. B. Conversion P rogram . . . . . . . . . . . . . . . . C. Proposed System . . . . . . . . . . . . . . . . . D. Orlando-Orange County Expressway Authority I mplications ............... ............. ........ IMPLEMENTATION ISSUES . .............. ......... A Technology . . . . . . . . . . . . . . . . . . .. B. Dedicated vs. Mixed Use Lanes ... .... ............... 1 2 5 5 7 9 12 13 13 22 C. Traffic Operations . . . . . . . . . . . . . . . . 29 D Accountability . . . . . . . . . . . . . . . . . . 31 E. L.egaJ Issues . . . . . . . . . . . . . . . . . . . 47 F. Computer System Requirements . . . . . . . . . . . 53 G. Ownership/Finance Options . . . . . . . . . . . . 59 SURVEY RESEARCH ... ...... ..................... A Survey of Existing A VI Use rs ........................ B. Other A VI-Related Surveys ........................ C. Florida Turnpike Patro n Surveys ..................... D. Market P otential .............................. ... COSTS AND BENEFITS . . . . . . . . . . . . . . . A Costs . . . . . . . . . . . . . . . . . . . . . B. Benefits . ................ .... .................. C. Cost Estim ates for A VI I mplem entation ......... .... . i 64 64 65 69 86 90 90 102 109


TABLE OF CONTENTS Section Page VII. COST-EFFECTIVENESS ANALYSIS OF AVI . . . . . . . 11 0 A. Introduction . . . . . . . . . . . . . . . . . . . no B. Methodology . . . . . . . . . . . . . . . . . . 111 C. Determination of User Benefits . . . . . . . . . . . Ill D. Determination of Costs . . . . . . . . . . . . . . 117 E. Conclusions . . . . . . . . . . . . . . . . . . . 121 VID. CONCLUSIONS AND RECOMMENDATIONS . . . . . . 127 APPENDIX . . . . . . . . . . . . . . . . . . . . 130 ii


LIST OF TABLES Table Page Table 1 Power Density Standards . . . . . . . . . . . . . . 20 Table 2 Rel ative C o mparison of A VI Te c hnologies . . . . . . . . 22 Tab l e 3 Volume 'Thresholds for A VI Implementat ion . . . . . . 28 Table 4 Top 10 Entry/Exit Points on Florida's Turnp i ke . . . . . 71 Tab l e 5 Question #1, Mail Back Survey . . . . . . . . . . . . 78 Table 6 Question #2, Mail-Back Survey . . . . . . . . . . . . 78 Table 7 Question #3, Mail-Back Survey . . . . . . . . . . . . 78 Table 8 Question #4, Mail Back Survey . . . . . . . . . . . . 79 Table 9 Question #5, Mail Bade Survey . . . . . . . . . . . . 79 Table 10 Quest ion #6, Mail-Back Surve y . . . . . . . . . . . 80 Table 11 Question #7, Mail-Back Survey . . . . . . . . . . . . 80 Table 12 Question #8, M ail -Ba c k Survey . . . . . . . . . . . . 80 Tab l e 13 Questi o n #9, Mail Back Survey . . . . . . . . . . . 81 Table 1 4 Que st ion #10, Mail-Back. Survey . . . . . . . . . . . 81 Table 15 Question #11, Mai l -Back Survey . . . . . . . . . . . 82 Table 16 Manual Lane Equipment Cost Est i mates . . . . . . . . 92 Table 17 Automatic Lane Equipment Cost Estimates . . . . . . . 93 Table 18 A VI Dedicated Lane Equ i pment Cost Estimates . . . . . 94 Tab l e 19 Mixed Use ( A VI & Manual) Lane Equ i pment Cost Estimates 95 Table 20 MixedUse (A VI & Manual) Lane Equipment Cost Est im ates . . . . . . . . . . . . . . . . . . 96 Table 21 P laza Computer Equipment Cost Estimates . . . . . . . 98 Table 22 Host C o mputer Equipment Cost Estimate s . . . . . . . . 100 T able 23 Total Road User Costs For No-Build and Improveme n t Alt ernatives . . . . . . . . . . . . . . . . . . . 114 Table 24 Road User Benefits, Peak Hour, Peak Direction . . . . . . 115 Table 25 Road Use r Benefits, Both Peak Hours . . . . . . . . . 116 Tab le 26 Lane Equ i pment Cost By Lane Type . . . . . . . . . . 118 Table 27 Number and Configuration of Lanes Under the Improvement Alternatives . . . . . . . . . . . . . 118 Table 28 Desirab ility Indicators Under Four Improvement Alternatives . . . . . . . . . . . . . . . . . . . 121 Table 29 Tamiami Plaza Benefit/Cost Analysis. Alternative 1: Construction . . . . . . . . . . . . . . . . . . . 123 Table 30 Tamiami Plaza Benefit/Cost Analysis, Alterna tive 2: 10% A VI Participati o n . . . . . . . . . . . . . . . 124 Table 31 Tamiami Plaza Benefit/Cost Anal ysis, Alternative 3: 30% AVI Participation . .. . .. . . . . . . . . . . .. 125 Table 32 Tamiami P l aza Benefit/Cos t Analysis, Alternative 4: 50% A VI Participa t ion . . . . . . . . . . . . . . . 126 iii


LIST OF FIGURES Figure Pag e Figure 1 Existing T oll Road Applications of A VI in the U nited States . . . . . . . . . . . . . . . . . . . 3 Figure 2 Pending Toll Road Applications of A VI in the United States . . . . . . . . . . . . . . . . . . . 3 Figure 3 Existing Turnpike System . . . . . . . . . . . . . . 6 Figure 4 Turnpike Conversion Program . . . . . . . . . . . . 8 Figure 5 Current and Scheduled Major Turnpike System Improvements . . . . . . . . . . . . . . . . . . 11 Figure 6 Average Capacity, Toll Plaza Lane Types . . . . . . . . 23 Figure 7 Average Speed, Toll Plaza Lane Types . . . . . . . . . 25 Figure 8 Plaza Lane Reduction Potential . . . . . . . . . . . . 26 F igure 9 A VI vs. Conventional, Toll Plaza Lanes . . . . . . . . . 27 Figure 10 Manual Lane Collection Process . . . . . . . . . . . 33 Figure 11 Automat i c Lane Collection Process . . . . . . . . . . 34 Figure 12 A VI (Express TOLL) Lane Collection Process . . . . . . . 35 Figure 13 TollTag Office Daily Cash Reconciliation . . . . . . . . 37 Figure 14 Nightly System Reports . . . . . . . . . . . . . . . 38 Figure 15 Ad Hoc Reports . . . . . . . . . . . . . . . . . 39 Figure 16 Special Action Reports . . . . . . . . . . . . . . . 40 Figure 17 Turnpike User Survey Locations . . . . . . . . . . . . 70 Figure 18 Turnpike Commuter Travel Patterns . . . . . . . . . . 72 Figure 19 Personal Int erview Resu lts . . . . . . . . . . . . . . 75 Figure 20 Percent of Job-Related Trips By Plaza . . . . . . . . . n Figure 21 Dual Plaza Computer Configuration . . . . . . . . . . 97 Figure 22 Host Computer System . . . . . . . . . . . . . . . 99 Figure 23 Vehicle Noise By Speed . . . . . . . . . . . . . . . 101 Figure 24 Fuel Consumption of Passenger Vehicles . . . . . . . . 103 Figure 25 Vehicle Emission Factors . . . . . . . . . . . . . . 105 Fi gure 26 Recommendations For A VI Implementation . . . . . . . 128 iv


I. INTRODUCTION AND OVERVIEW The Florida Department of Transponation bas contracted with the Center for Urban Transportation Research (CUTR) at the Uni versity of South Florida (USF) to perform an "Analysis of Automatic Vehicle Identification and Its Potential Application on the Florida Turnpike." The project is comprised of three distinct but related phases. Phase I consisted of a review of the state of the art in automatic vehicle identification (A VI) technology including a survey of vendors and operators of A VI systems to determine available technologies, system operating characteristics, and examples of applications nationwide. The information gathered in Phase I was compiled and presented in Technical Memorandum 1. A synopsis of this memorandum is provided in Section II of this report. Phase II consists of survey research regardin g attitudes a nd characteristics of e xisting A VI users as well as current Florida Thrnpike patrons. This phase includes five survey research efforts including surveys of existing A VI users, summary of other A VI-related surveys, and Florida Thmpike patron interviews, mail -back surveys, and focus groups. Phase ill of the project involves an evaluation of the specific application of A VI technology to Florida's Thrnpike The evaluation includes a m ore-deta il ed consideration of the issues defined in the first technical memorandum including A VI technology, payment systems, dedicated vs. mixed-use lanes, traffic operations and safety, speed vs. accountability, enforcement, legal i ssues, a nd others. Additional issues have also been identified and are co nsidered in this final phase. In addition, significant use will be made of Phase II survey efforts to assess the market potential and anticipated level of participation in an A VI system on Florida's Turnpike. After carefully considering the issues and survey results, the specific costs and benefits of various components of an A VI system will be defined and q uantified where possible. The results of Phase II and Phase ill efforts are summarized in this technical memorandum. 1


II. SYNOPSIS OF TECHNICAL MEMORANDUM 1 The efforts of Phase I of the project were presented in the report entitled "Analysis of Automatic Vehicle Identification Technology and Its Potential Application on the Florida Turnpike, Technical Memorandum 1." This memorandum included an overview of general applications of automatic vehicle identification (A VI), a review of A VI technology as applied to toll collection, existing toll road applications of A VI, pending toll road applications of A VI, and issues related to the application of A VI. Included in an appendix were a survey letter to A VI vendors, a list of A VI vendors, a survey letter to A VI operators, a list of A VI operators, and a bibliography related to A VI technology. A VI originated in the railway industry, with the purpose of tracking the movements of trains and contributing to more efficient scheduling. The concept has also been used by the shipping industry for cargo tracking and port of entry clearance. The trucking industry has also used A VI for vehicle tracking and weigh station processing. Finally, A VI technology is rapidly evolving in the area of toll collection. Automatic vehicle identification in the toll collection industry refers to techniques that uniquely identify vehicles as they pass specific points along a facility, without requiring any action by the driver or an observer. For this application, there are four major technologies currently being developed and implemented around the world: optical/infrared (bar code), inductive loop, radio freque ncy (RF)/rnicrowave, and surface acoustical wave (SAW). The general characteristics of these technologies were reviewed and included in Technical Memorandum 1. In addition to reviewing the state of the art in technology, the report also reviewed current A VI configurations as well as various characteristics of administration including ownership arrangements and payment systems. Numerous existing toll road applications of A VI were reviewed, each of which included a brief description of the system, the technology employed, the method of payment for A VI users, and other key issues. Systems included in this review are presented in Figure 1. Also, numerous pending toll road applications of A VI were summarized including those represented in Figure 2. 2


FIGURE 1 Existing Toll Road Applications ot A VI i n the United States Callao North v .. ...,ollway FIGURE2 Say Harbor be County "J Pending Toll Road Application s of A VI In the United States Oran ge '"-Q!!fl Calirornr 3 Y-NJ-<;ONN :-\ County Rieken backer Causeway Puerto Rico


Finally, various issues related to the application of A VI were identified and briefly reviewed as part of Phase I. Iss ues identified included market identification, benefits and costs, dedicated vs. mixed-use lanes, traffic safety issues, speed vs. accountability, legal considerations, payment systems, potential health effects, and electrical in terference Other issues have been identified and will be addressed in this technical memorandum. 4


III. DESCRIPTION OF FLORIDA TURNPIKE SYSTEM EXISTING SYSTEM The existing Florida Turnpike system includes a "mainline," the Homestead Extens ion of the Florida Turnpike (HEFT), and the Bee Line West Expressway. In Figure 3, the Turnpike's 266-mile mainline runs from Interstate 75 in Sumter County to the Golden Glades exit in Dade County. This section of the Turnpike is designed for long-distance trips as well as commuter and business trips in the densely populated areas in the southeastern region of the state. The HEFf is a 47-mile expressway extending west from the mainline near the Dade/Broward County line southward to U.S. Highway 1 (see Figure 3). The HEFf serves the rapidly developing Dade County area west and southwest of Miami. In Figure 3, the Bee Line West, an 8-mile east-west connector in Orange County, extends from Orlando International Airport westward to Interstate 4 near Walt Disney World, with an in termediate inte rchange at Orlando.South on the mainline The existing Florida Turnpike system is a total of 321 miles in length and presently includes an estimated 204 toll lanes. About 120 toll lanes exist within the primary urban area of the Turnpike system south of the Lake Worth interchange. Approximately 15 toll lanes exist within the secondary urban area of the Turnpike system between Orlando West and the Kissim mee/St. Cloud interchange The rema ini ng 69 toll lanes are located along the vast rural portions of the Turnp ike system. 5




CONVERSION PROGRAM The Florida Department of Transportation is implementing a three-stage toll rate increase and collection system conversion plan designed to improve toll collection along the Turnpike. Figure 4 illustrates the limits (overlapping in some instances) of this staged conversion program. In Stage 1, which was recently completed, tolls were increased approximately 50 percent in 1989, and the present ticket system of toll collection on the mainline in Dade County, Broward County, and the southern portion of Palm Beach County was converted to a coin system on August 19, 1990. In Stage 2, scheduled for implementation on July 1, 1991, tolls will increase 25 percent from the current rate for that portion of the Turnpike system north of Stage 1 (Lantana). Stage 3A will increase the tolls another 25 percent across the entire Turnpike system on July 1, 1993. Finally, in July 1995, Stage 38 will convert the present ticket system of toll collection to a coin system on the northernmost end of the Turnpike (future Poinciana interchange to Wildwood) and add new ramp toll plazas at the southernmost HEFT interchanges and westernmost Bee Line interchanges to "close up" the system. 7


qulf o/ 'Mexjco F I GURE 4 T U RNPIKE CONVERSIO N PROGRA M STAGE 38 BEE UNE W EST __. EXPRESSWAY II. y CT ORL ANDO STAGE 1 completed Augu st19, 1 990 Ill Ill I I STAGE 2 completion Jul y 1, 1991 STAGE 3A completion Jul y 1, 1993 STAGE 38 completion July 1, 1995 FUTURE POINCIANNA INTERCHANGE NOTE: STAGE 2: Norlh oll.anlana Rd. will onl y I ncrease l olls b y 25%. STAGE 3A: Entir e Turn p ike SySiem will only increase lolls by 25%. STAGE 2 FLORIDA'S TURNPIKE .. MaJn Slteet '\&A \ _IIIOWI R!V CO. Sf. lUCECO. LANTANA RD ..., f>J. llot WtCH 00..- 8'QWARO C(), STAGE 1 STAGE 38 HOMESTEAD () .91t fantic Ocean WEST PAlM BEACH FORT LAUDERDALE M I AMI w-\>- 0 5 10 20 ml


PROPOSED SYSTEM The Florida Department of Transportation has planned additional Turnpike improvements, not all related to toll rates and toU collection systems, through fiscal year 2000. These projects will improve the Turnpike facility and enable it to continue to provide a high level of service. The principal improvement affecting Turnp ike capacity is the addition of toll lanes and the widening of the Turnpike from four to six lanes between Golden Glades in Dade County to Boca Raton in Palm Beach County. The FOOT has also scheduled additional widening, interchanges, and minor improvements on the Turnpike through 1995. The following is a list of all scheduled Turnpike improvements. WIDENING PROJECTS 6 lanes fro01 SUDrise to Oaldand, 1988-1989 6 lanes from Commercial to C-14 Canal, 1988-1989 6 lanes from Golden Glades to Mirimar, !988-1989 6 lanes from Government Center to Coral Reef Drive, 1988-1989 &lanes from Coral Reef Drive to South Dade Expressway, 1988-1989 Kendall Drive to Tamiami Trail, 1988-1989 61anes from Pompano Beach to Golden Glades, 1988-1989 Glades Road bridge, 1988-1989 6 lanes from C-14 Canal to Sample Road, 1989-1990 Okeechobee Blvd. from SR-7 to Haverhill Road, 1989-1990 Lake County bridges, 1989-1990 S umter County bridges, 1990-1991 INTERCHANGES Dolphin Stadium, 1988-1989 N.W. 57th Avenue at Broward County Line, 1988-1989 Griffin Road, 1988-1989 lnterstate-595 (fully directional, no tolls)under construction Boynton Road in Palll1 Beach County, 1989-1990 S.W. 110th Street in Dade County, 1989-1990 N.W. 41st Street in Dade County, 1990-1991 N.W. 1061h Street in Dade County, 1991-1992 Atlantic Blvd. in Broward County, 1992-1993 Dart Blvd. (Poincianna Interchange) in Osceola County, 1992-1993 Southern connector in Orange County, 1993-1994 9


ROADWAYS Ramp$ at HEFT/Mainline Turnpike, 1 990-1991 N.W Hillsbo r ough Expressway, 1991-1992 Seminole County Expressway SR 426to US 17-92, 1991-1992 lOth Stree t extension of Sawgrass, 1993-1994 OTHER PROJECTS Toll E q uipmeDI Improvements 1 990-1991 Convert to Coin System at Lantana, 1990-1991 Resurf acing, 1990 1992 Impr ove Lake Worth Road at Turnpike, 1991-1992 Pave Median Shoulders, 1991 Coovertlo Coin System oorlb of caooe Creek, 1 994-1995 Figure 5 ill ustrates the location of major current and scheduled Turnpike system improveme nts. 10



OrlandO=Oran&e County ExpresswaY Authority lmo!icatjons Originally, the Orlando-Orange County Expressway Authority (OOCEA) issued a request for proposal (RFP) for A VI system equipment. The RFP was expanded to include manual fare collection and full A VI integra tion equipment in the bid specification. Bids were received on June 14,1990. Due to the inconsistency of the technical and cost proposals, the OOCEA Board of Directors rejected the June bids and will revise the original specification to include a more specific system configuration and approach. The second stage of bidding is expected to take place by the beginning of October 1990 with proposals due approximately January 1, 1991. Unit prices from the June bids were reviewed and found to be comparab l e to costs utilized in the cost effectiveness analysis of this report, with only minor discrepancies. Another aspect of this project is also worth noting: full A VI integration was originally specified, whereas the revised specification requires the installation of A VI on one plaza and the monitoring of results before further commitment. This stagi ng approach to A VI appears to be more practical and more cost-effective, especially considering the assurance needed for technology performance and software integration. Also, it is appropriate that the Florida Turnpike and the OOCEA coordinate efforts on performance/acceptance testing of A VI technologies. 12


IV. IMPLEMENTATION ISSUES Numerous implementation issues were identified and reviewed including: technology, dedicated vs. mixed-use lanes, traffic operations, accountability, legal issues computer system requi rements, and ownership/finance options. TECHNOLOGY All present A VI technologies operate by (1) intercepting modulated electromagnetic radiation from a veh i cle, (2) recovering the i nformation contained in the signal, and (3) using a computer to identify the vehicle from a database. The differences i n the various technologies are in the ways intercepting and recover i ng are accomplished. Technologies may be divided according to the frequency of the electromagnetic radiation, the method by which the signal is modulated, and whether the vehicle tag generates or simply reflects electromagnetic radiation. There are three frequency ranges in use: (1) very low frequencies (below 200kHz), which are employed in inductively coupled systems; (2) microwave frequencies (500 to 3000MHz); and (3) optical and near-optical frequencies, which include infrared. lnductjve LooP Systems The only A VI technology that employs very l ow frequencies is the inductively coupled SYStem, which uses a loop antenna imbedded beneath the surface of the roadway to communicate with a tag mounted on the underside of the vehicle The roadway antenna sends out an interrogation signal and the tag responds by returning a signal which is modu lated according to data stored in the tag. This is normally an active (as opposed to a passive) system since the tag normally transmits its own signal (rather than reflecting the interrogation signal). This is the oldest of all the A VI technologies and, from a theoretical point of view should be one of the mos t reliable due to the very short distances between the imbedded antenna and the vehicle tag. 13


The advantages of this type of system are: (1) potential for greater reliability due to close proximity of loop antenna and tag; (2) simple serviceability; (3) very low potential for electrical interference; and ( 4) low potential for interference from adjacent lanes due to short coupling range The disadvantages of an inductively coupled system are: (1) low frequency resulting in lower maximum data rate, although it i s fast enough to allow multiple transmissions to increase reliability; (2) larger and more complex vehicle tags; (3) medium difficulty in duplicating tags; (4) tag usually requires power from vehicle (active tag); and (5) tag installation is more difficult. Optical Systems There are rwo basic types of A VI technologies that employ optical or near-optical frequencies to identify vehicles The first is a system that reads license plates directly and identifies the vehicle from a database. As the vehicle passes the tollbooth, a video camera forms an image which is digitized and processed to extract the license plate number. Typically, the image processing can take nearly one second so that multiple reads to improve reliability are not possible The advantages of this type of system are: (1) no special vehicle tag is needed; (2) license plates are not likely to be duplicated; and (3) there is no chance of interference berween adjacent l anes. 14


The disadvantages of this type of system are: (1) the processing equipment and algorithms are very complex; (2) the relatively long time required for image processing precludes multiple reads to increase reliability; (3) the system is subject to failure due to dirty or damaged license plates the presence of bumper stickers and similar text on a vehicle, and reduction of visibility due to rain and fog; and ( 4) very low (80-90%) reliability due to the complexity involved in image processing. The second type of optical or near-optical system employs a vehicle tag that is simply a bar code. A laser scans continuously over the area where the tag is expected to be and the reflected signal is processed to extract the code This image processing is much simpler than trying to read a license plate, since the reflected laser signal represents a one dimensional image whereas the video image of the license plate must be processed in two dimensions The advantages of this type of system are: (1) greater reliability than systems reading license plates due to the single dimension; (2) very simple vehicle tag which is just a bar code imprinted on a plastic card; (3) low potential for lane to lane interference due to limited range; and (4) much faster than systems which read license plates. The disadvantages of bar code systems are: (1) tags are easier to duplicate compared to other A VI technologies; (2) susceptibility to failure caused by rain, fog, and dirt or moisture on tag; and (3) necessity of image processing for finding the returned signal results in less reliability than systems employing transponders (microwave systems). 15


RF /Microwave Systems Radio-frequency (RF) A VI systems employ microwave frequencies to communicate to and from the vehicle. All microwave systems have high data rates which allow multiple transmissions (redundancy), resulting in increased reliability. These multiple transmissions are commonly known as "handshakes" in the industry. These systems may be divided into those in which the tag generates its own microwave signal (active tag) and those in which the tag simply reflects the microwave signal that it receives (passive tag). Active tags require a power source (battery or connection to vehicle power) while passive tags may or may not require a power source. Microwave systems may also be divided according to the method used to store data in the tag. The codes that identify a vehicle may be stored in an integrated circuit memory or in a "'surface acoustical wave" device (SAW). An integrated circuit memory may have a fixed code or may be programmable, while a SAW device is manufactured with a code which cannot be changed. In an active vehicle tag system, the transmitter at the tollbooth sends out a very short interrogation signal which triggers the circuitry in the tag. The tag responds by generating a microwave signal that is modulated with the data stored in the tag. This signal is transmitted to a receiver at the tollbooth which decodes the data and sends it to a computer for identification. The advantages of an active vehicle tag system are: (1) greater operating range than a passive system since the tag is not powered by the interrogating beam; (2) greater reliability than a passive system since the return signal from the vehicle is much stronger; and (3) less chance of electrical interference since the signals are stronger The disadvantages of an active vehicle tag system are: (1) greater complexity in the tag circuitry; (2) greater probability of lane to lane interference due to stronger signal; and (3) the tag must have a battery or be connected to vehicle power. 16


In a system that employs a passive vehicle tag, the transmitter at the tollbooth must transmit a signal continuous ly. This signal is intercepted by the tag and reflected to a receiver at the tollbooth The amount of reflection is varied (th e reflected signal is modulated) according to the data stored in the tag. The received signal is decoded to recover the data, which is sent to a computer fo r id e ntification The advantages of a vehicle tag system are: (1) the tag does not need a battery to be connected to vehicle power; (2) t he tag is less complex than in an active system ; and (3) there is less chance of lane to lane interference due to the lower signal power levels. The disadvantages of a passive vehicle tag system are: (1) lower reliability than an active system; (2) greater suscept i bility to electrical interference due to lower signal levels; (3) shorter operating range since tag is powered by the interrogating beam; and (4) the overall level of microwave radiation is higher The advantages of using an integrated circuit memory to store the vehicle tag data are: (1) the tag is reprogrammable; and (2) less sophistica ted equ i pment is required for tag manufacruring. The disadvantages of an integrated circuit memory are; (1) the tag circuitry is more complex; and (2) only medium difficulty in duplicating tags. The advantages of using a SAW device to store data in the vehicle tag are: (1) it is virtually impossible to duplicate the vehicle tag; and (2) t he tag circuitry is much simp ler. 17


The disadvantage of using a SAW device is: ( 1) a limited operating range (up to 15 feet) since it is normally part of a passive system. Electrical Interference The possibility of a failure to properly read a vehicle tag due to the presence of an unacceptable level of electrical interference must be considered when implementing an A VI system. Electrical interference can occur in an A VI system in two different ways. The first is due to other (non-A VI) transmitters operating on the same or nearby frequencies and are close enough to produce a strong interfering signal. Possible sour ces of this type of interference are cellular telephones, police and other mobile communications, and radars. The possibility of this type of interference can be minimized by obtaining an FCC license for an A VI system. This accomplishes two things. First, a larger transmitted power level may be used, and second, a frequency will be assigned that is different from other radio services operating in the same area. An unlicensed system can only depend on redundancy of transmission to reduce interference and, according to the FCC, "must accept any interference that may be received including in terference that may cause undesired operation." The second type of interference is that which can arise in the A VI system due to improper design and/ or installation. This can occur if the transmitted signal from one A VI lane of traffic is allowed to enter another A VI lane and can result in multiple recordings of the same vehicle or failure to record one or several vehicles. The remedy for this type of interference is proper design of the A VI system and, in particular, proper selection and placement of all antennas. Health & Safety Factors The rapid increase in the use of electromagnetic energy in modem soc iety has led to an increased public awareness and concern regarding possible (or perceived) effects on human 18


health. Use of the word "radiation" causes fear in many people and, since radio frequency waves and microwaves are "radiation," their potential use in an A VI system must be addressed from the point of view of both actual and perceived hazards. The effects of radiation may be divided into three categories: ionization, heating effects, and biological effects. Of these, only the latter is of importance in A VI systems. Ionization can occur only for very high frequencies such as x-rays and never occurs at radio and microwave frequencies. Heating effects (such as in a microwave oven) occur only for very high power densities and are of no consequence for the power levels used in A VI systems. The third category, that of possible biological effects, is the one that must be addressed prior to the implementation of an A VI system. The probability of harmful biological effects is a function of both the frequency and power density of electromagnetic radiation. The frequencies used in A VI systems extend from very low to about 3000MHz, which is in the microwave regi on. Generally, greater biological effects are thought to be associated with higher frequencies since these have higher energies pe r photon of radiation. Therefore, the higher frequencies (800 to 3000MHz) will be eonsidered in the analysis which follows. An even more important factor in determining the likelihood of harmful biological effects is the power density associated with radiation This is a function of both the power level of the transmitter and the type of antenna used. Antennas concentrate the power (similar to a lens) and the amount of eoncentration is specified by a term called "antenna gain." The type of antenna that produces the greatest eoncentration of power is the parabolic dish antenna. A 4-foot diameter, parabolic antenna has a gain of approximately 20<18 at 900MHz and will be assumed in the discussion that follows. A gain of 20dB means that the power is concentrated by a factor of 10. If we assume that the power output of the A VI transmitter is 500m W, then the maximum power density at a distance of o ne meter from the antenna can be calculated from standard antenna theory to be approximately 40p. WI cm2 This is the power density that should be eompared to nationally accepted safety standards. A summary of the various standards from relevant agencies is shown in Table 1. 19


TABLE 1 Power Density Standards American National Standards Institute ANSI C9S.l 3 ,0CXJ..,.w /em' lntem.ational Electroteclmical Commis:sion w /em' National Radiological Proteetioo Board fon2 Occupational Safety & Health Agency OSHA 1910 .97 Jon' Environmeotal Pr01ectioo Agency (proposed standard) Soviet Occupational Standard (2 hours exposure per day) As can be seen from the above data, the powe{ densities encountered in A VI syste{DS are far below all accepted ( 0{ proposed) national standards. When one also conside{S the facts that the transmitter power level will probably be less than SOOm W, that the A VI user will be more than one meter from the antenna, and that the exposure time will be less than one second, the A VI system must be considered to be one of the safest applications of electromagnetic energy. To put the A VI power densities into proper perspective, one need only consider an increasingly common modem convenience the portable cellular telephone. A typical ponable cellular telephone transmits 4 watts at a distance of about 5 centimete{S from the operator. The associated power density is about 6,4001' W or about 160 t imes that of the A VI system. Metal Oxide Windshields Some 1991 luxury ca{S are being equipped with a metal oxide coating on the windshield. This coating reduces solar radiation by 30 percent and ultraviolet radiation by 45 percent, resulting in improved air conditioner performance and prolonged interior material life. This metal oxide coating causes disruptions in A VI signals that must pass through the windshield. At this time, it is not certain how widespread the use of metal oxide will be; however, it is an important implementation issue to consider. It is expected that the disruptions caused by metal oxide could be avoided by adjusting the placement of the transponder/tag. 20


Tae Installation The reliability of A VI tags can be assured if the tag is permanently installed. Vendors prefer, and sometimes require, that tags be permanently mounte d by experienced toll agency personnel for this reason. However, for patron convenience, some tags are installed such that t hey can be removed. This convenience may cause "misreads" when tags are not placed in the expected location. Recommendations Hard results from techno logical performance testing does not appear to be available and, where problems have occurred, it is difficult to determine if the problems result from the reliability of the technology or from its application. However based on information provided by vendors, visits to various A VI sites, and general discussions with recognized industry experts, relative comparisons among technologies can be offered. Seven major technology issues can be identified to be used in relative comparisons. A relative comparison score of high, medium, or low can be attributed to each technology regarding each issue. It is importan t to note that the specific differences between high, medium, and low are not known at this t ime, only that there is a relative difference. "High" is most favorable and "low" is least favorable. Although the issues have been identified, they have not been weighted according to their relative importance. Turnpike officials will need to establish the relative importance of the identifie d issues in order to select a technology In summary, Table 2 provides the relative comparisons. 21


TABLE2 Relative Comparison of A VI Technologies ISSUES/TECHNOLOGIES RF /Microwave SAW loductlve Loop Bar eo.u Reliability high medium high low Reslslallce to Dupllcatloo medium high medium low (security) Poteotlal Cor Muldple Reads high high low low (speed vs. reliability) Reslstaoce to lnterCereoce low low high high (lane-to-laue) Tolerance to Environment high high medium low Simplicity or Tag (cost) low medium low high Health Safety high high high high DEDICATED vs. MIXED-USE LANES The issue of dedicated vs. mixed-use lanes for A VI implementation is dependent on four basic characteristics: capacity by lane type, the relationship of speed to capacity, levels of A VI participation, and thresholds for toll plaza lane configurations This section will discuss each characteristic and how they are interrelated. Capacity bJ' Lane 'IJl!e Existing and future toll plaza lanes can be characterized into five basic lane types: manned, automatic mixed A VI, dedicated A VI (within a conventional plaza), and express A VI. Manned toll lanes require all toll transactions to be handled by a toll collector. Automatic lanes collect tolls by providing coin machines (in Florida these types of lanes are also gate controlled). Mixed A VI lanes combine A VI with either manual or automatic toll collection. Dedicated A VI lanes are contained within conventional toll plazas but permit A VI patrons 22


only. Express A VI l anes are physically separated from all other type toll lan es permitting free -flow (SSmph or greater) speeds. Capacities for manned, automatic, and mixed A VI lanes were determined from observations and counts from existing toll facilities such as the Florida Turnpike, Tampa Crosstown Expressway, the New Jersey Turnpike, and the Dallas North Tollway. Capacities for automatic lanes were determined from Florida Turnpike automatic lanes which all have gates. N o other capacities were determined for lane types with gates. Average capacities are typically reduced by 10-20 percent when gates are used on automatic lanes; the higher capacity effect due to increased gate sensitivity is not known at this time. Capacities for mixed-use A VI lanes were obtained from observations on the Dallas North Tollway. Mixed A VI lanes at this site include A VI on both manual and automatic lanes (with no gates). Capacities for dedicated A VI and express A VI lanes were estimated based on reasonable average-speeds and vehicle spacing (i.e., headways). Figure 6 illustrates the general relationship of average capacity, or throughput, for the basic types of toll plaza lanes. Depending on plaza lan e configuration, the inclusion of A VI has the potential to increase conventional plaza lane capacity by 50-160 percent. FIGURE 6 Average Capacity, Toll Plaza Lane Types CAPACITY (VEHICLES/HOUR) 2,000 1,500 ----.. -. --------1 ,000 500 500 350 0 MANNED AUTOMATIC MIXED .WI OEDICATii:O Jill EXPAESIJ 14/1 23


Speed-Capacity Rela t ionship The average capacity for each plaza lane type is a factor of the speed in which vehicles proceed along the queue through the toll plaza Speed is based on the rela t ionship of volume divided by the density. The greater the density (or as the spacing between vehicles becomes smaller), the lower the speed becomes, resulting in a lower throughput volume across a fixed point over a given time period. Conversely, as the density decreases (or as the spacing between vehicles becomes greater), the speed increases and thus a higher throughput volume can be realized across a fixed point over a given time period. Safe and reasonable vehicle spacing for various speeds has been previously determined from established traffic engineering theory and observations. For example typical spacing for 10 mph is taken as 55 feet (center t o center of vehicle), and for 55 mph, vehicle spacing is taken as 160 feet At lOmph, with 55-foot s pacing of vehicles, density would be 96 vehicles per mile and the corresponding volume passing a fLXed point in an hour would be 960 vehicles/hour. At 55mpb, with 160-foot spacing of vehicles, density would be 33 vehicles per mile and the corresponding volume passing a fiXed point in an hour would be 1,815 vehicles/hour These respective throughpu t volumes are assumed to be the capacities for the applicab l e lane types. Figure 7 depicts the relationship of speed to the various toll plaza lane types. As can be observed from this figure, speeds (and volumes) increase as preferential treatment for A VI increases because there is l ess restriction in t oll processing. For example, ave r age running speed (the speed maintained once a vehicle first stops or slows down while approaching the toll lane queue through the point of being processed at the toll plaza) for a manual lane is about 2.5mph, compared to approximately 55 mph for an express A VI lane Higher average speeds through the utilization of A VI lanes translates into greater lane capacity potential. However, with greater speeds, A VI lanes retrofitted into a conventional toll plaza will require additional planning, design, and right of-way (as in the case of express A VI lanes). 24


FIGURE 7 Average Speed, Toll Plaza Lane Types SPEED (MILES/HOUR) 70 60 50 40 30 20 10 2.5 0 MANNI!O AUTOMATIC toUXI:O .WI Oli!OICATIEO ,IQ{I EXPAI!$8 .WI l&vels of AVI Participation Actual patronage levels for A VI are extremely difficult to estimate. Therefore, full utilization of toll lanes that are retrofitted with A VI or physically separated from the conventional plaza for express A VI usage can only be assumed, given an estimated level of A VI participation. The highest current A VI participation rates being experienced are on the Treasure Island Causeway ( 60 percent), Bay Harbour Island Causeway ( 40 percent), and Dallas North Tollway (23 percent). These levels of A VI participation do not come without some additional cost for marketing or publicity, such as the professionally operated, high profile tag purchasing store in Dallas. The effect of varying A VI participation levels on plaza lane requirements is important to distinguish. The Florida Department of Transportation Traffic Engineering Office has recently developed the first generation of a bas ic 1011 plaza simulation model known as D QUEUE. A basic validation of this model was performed successfully with results matching observed conditions at three locations: Tampa Crosstown Expressway at East Plaza, HEFT at Tamiarni Plaza, and Commercial Boulevard exit plaza. DQUEUE was utilized in this 25


feasibility study to approximate the potential for plaza lane reduction utilizing A VI under various participation rates for various peak-hour volumes. Maximum allowable queues, used as the criteria for determination of ideal plaza lane configurations, were taken to be no more than 300 f eet (FOOT Turnpike planning standard). Figure 8 displays the results of over 1 00 D-QUEUE model runs to determine the maximum potential for toll plaza lane reduction for peak directional volumes of 3,000-6,000 vehicles per hour The parameters for t hese D-QUEUE runs are also i ncluded in the appendix of this report. Runs were made for conventional lanes only and for conventional lanes with A VI to discern the maximum potential for plaza lane reduction at similar approach volume l evels It can be clearly observed from this figure that, as volumes (plaza lanes) increase, the potential to reduce the n umber of toll plaza lanes is enhanced by a greater A VI participation rate FI G URE 8 Plaza Lane Reductio n Potential Lante Reduced CEach Direction) e,ooo I 8 --. ----... --::;;; __.. ...-.c:::._::_ .,_" ---5,000 4 ,000 3,000 4 2 0 0'10 10'10 20'10 30'10 40'10 50'10 60% 70% 80% 90% 100% 3000 vph 5000 vph Greater Lane Reduction Pot enllal Oeeura af P lazas Wltl'l Hlghttt Volurntt/Number of L anee PeakHr,Peak-Dir Vo l 4000 vph 6000 vph 26


Thresholds fgr Toll Plaza Lane O!nfiamratigns Combining the previous fin dings of capacity by lane type (see Figure 6), and the relationsh i p of the potential for lane reduction with various A VI participation rates at different volumes (see Figure 8), ideal toll plaza configurations can be determined Ideal configurations assume the fewest and most practical number of toll plaza lanes so as not to exceed the maximum allowable queue length of 300 feet. The D-QUEUE toll plaza simulator model was again utilized to determine t he fewest number of lanes for both conventional-only and conventional with A VI configurations. Figure 9, which builds from the relationship indicated in Figure 8, indicates the results of over 100 D-QUEUE model runs that were performed to determine ideal configurations FIGURE 9 A VI vs. Conventional, Toll Plaza Lanes F.,_at Nca. or Toll Plua Lan .. (Bolli Olrtctlona) 30 25 20 15 10 5 o-== ,, --1 ,000 2.000 3,000 4,000 5,000 6,000 7,000 Volume Pet' Hour [ AVI B Conventional I M in, Lan e Aeoulrementa Bd on 300tt Ma. Oueue, at \.e aat 20 Perctnt /NI PerUclpaCion For example, at 3,000 vehicl es per hour a conventional-only lane plaza would require a minimum of seven lanes in each direction (5 automatic and 2 manned). The most practical pote ntial for lane reduction would require about a 25 percent A VI participation rate to reduce th e number of lanes to five lanes in each direction ( 1 manned and 4 mixed A VI). A 25 percent A VI participation rate wo uld equal about 750 A VI patrons in the peak-hour, p eak-d irect ion, and four mixed A VI lanes would provide an additional capacity of 80 0 vehicles/hour (200 vehicles/hour /lane increase in capacity between automatic and mixed 27


A VI lanes). The results indicated in Figure 9 follow this same rationale for determining the fewest number of toll plaza lan es in each direction. Several key thresholds can be deduced from the examination of Figure 9. At 2,000 vehicles per hour, the A VI and conventional lane requirements are about the same. At 3,000 vehicles per hour, a significant difference in lane requirements becomes apparent indicating the initial threshold for A VI consideration (i.e., mixed-use A VI). A similar type of lane reduction pattern exists again at 5,000 and 7,000 vehicles per hour suggesting more intensive use of A VI. These thresholds are developed under ideal conditions, but nevertheless can be used as planning guidelines. Table 3 provides the planning guidelines for A VI implementation. TABLE3 Volume Thresholds ror A VI Implementation Initial Consider.>tion for Mixed A VI 3,000 vehicles per hour Initial Consideration for Dedicated A VI 5 000 vehicles per hour Initial Consideration for Express A VI 7,000 vehicles per hour Recommendations The issue of dedicated vs. mixed-use toll plaza lanes is complex and must be considered on a sitespecific basis. An integral aspect of this issue is the expected level of A VI participation and thus the number and configuration of lanes required. The commitment to A VI in terms of patron satisfaction does not necessarily require express A VI lanes (although this feature is most appealing based on the Florida Turnpike Use r Survey). Based on the analysis summarized above, under ideal conditions, expected future (year 2015) volumes on Florida's Turnpike will not warrant any higher preferential treatment for A VI than dedicated lanes within a conventional toll plaza configuration since the future peak-hour, peak-directional volumes (at Tamiarni mainline barrier) are not expected to exceed 7,000 vehicles/hour. zs


TRAFFIC OPERATIONS Traffic operations with and without A VI will differ significantly. In particular, current conventional treatments and methods for vehicle classification, toll collector safety, gates, and signing/ channelization will need to be re-evaluated. Vehicle Classification Veh icle classification is currently performed as a two-part process. Typically, the first part of this process consists of the toll collector visually classifying an oncoming vehicle to determine fare requirements (pre-classification). Second, treadles, which are used to count axles, are utilized in the post-classification process to verify t he manual vehicle classification. Inefficiencies result because toll collectors may not accurately determine the vehicle classification and treadles can frequently malfunction due to the wear of mechanical contact switches over time particularly in high-volume, high-truck lanes. Improved treadles and automatic vehicle classification systems already being considered as part of the Turnpike's toll collection system upgrade will dramatically increase the accuracy . of vehicle classification. Electronic toll collection, or A VI, will also reduce toll collector errors and fraud on A VI transactions by providing the audit trail for each A VI-equipped vehicle that is processed through a toll plaza lane. Toll Collector Safety Any implementation of A VI within a conventional plaza or a separated facility will present a safety problem for toll collectors when arriving a nd departing from their toll booths. As mentioned previously, with the implementation of A VI comes an increase in average speeds of vehicles through the toll plaza Most of Florida's Turnpike plazas do not provide pedestrian tunnels or overhead walkways and thus consideration for this f eature should be given at A VI-retrofitted plazas as well as separated plazas where safe access may be required for maintenance purposes. 29


Gates Gates are currently used in automatic lanes o n Florida's Turnpike as a means of deterring vio l ators These gates add one to one-and-one half seconds to the processing time of each transaction, thereby reducing capacity or throughput. Gates can also malfunction and create backups. Gate arms can also been broken or snapped off by vehicles passing through the plaza, which also results in delays According to the Florida Thmpike User Survey, gates would not be desirable with A VI implementation However, if state statutes cannot be modified to allow for photographic enforcement, then gates may be the only answer to deterring violators. Photographic enforcement will be considered more closely later in this report Advance S i p i n e /Chan n elization Procedures have been established by the Florida Turnpike regarding traffic control (signing, pavement striping and channelization) and traffic rules (speed and passing zones). These procedures represent criteria that apply to conventional toll plaza configurations and are intended to provide for the safety of the toll road patron. As various configurations of A VI intensity are implemented (mixed use lanes to dedicated to express) more attentio n will be required for advance sig ning and channelization to safely accommodate A V I. Site-specific speed limits and transition requirements to diverge and merge A VI traffic have to be developed and established as vital design criteria. Rttommendations Current traffic operations should be revised as necessary with the implementation of A VI. Vehicle classification procedures and toll collector safety will be i mproved even with the planned upgrade of toll lane equipment. New laser scanners for vehicle classification and tunnels for the movement of toll collectors could be worthwhile components of an A VI system. The use of gates and revised advance signing/ channelization needs further evaluation. The most efficient, cost-effective means of A VI enforcement and vehicle guidance must be developed. 30


ACCOUNTABILITY Operational and Accountine Functions The ability to effectively track toll revenue, process patron account information, generate relevant management and operational reports, and safeguard toll system assets is an important goal of any A VI system. All of these operational and accounting functions are discussed in the following sections, as they relate directly to A VI toll collection systems. Trackine Toll Revenue -The ability to track (audit) toll revenues is an imponant function of a n y toll facility. Figures 10 through 12 give an overview of manual, automatic, and express A VI lan e collection processes from when toll revenue is collected until it is deposited. These diagrams were provided by the E-470 Public Highway Authority in Denver, Colorado. First, in the manual lane coUectjon process diagram (Figure 10), the physical flow of money is detailed. The flow begins when the toU collector's shift ends: tolls coUected during the shift are counted and the amount is recorded on a deposit sheet After this, the collector's supervisor verifies the correctness of the deposit sheet and then deposits the tolls into the money room. A designated employee other than the collector or coUector's supervisor counts the tolls again to verify and record the deposit At this time, the tolls are soned, consolidated, and sealed in preparation for the bonded carrier to retrieve. The bonded carrier then takes the toll money to the bank where the bank recounts and verifies the deposited amount. In the next section of the d iagram, the data processing steps are detailed to coincide with the physical flow of money. Within data processing, the violation cameras are used to record lane violations while surveillance cameras are used to observe toll agency personnel. Also, the lane controller is directly linked to the host computer via the plaza computer. This enables the flow of transaction information to and from the various databases. The lane controUer is also directly linked to the terminal screen, which allows the toll coUector to see information concerning each A VI transaction. In addition, the treadles and vehicle detection loops give relevant axle and vehicle counts to tbe lane controller. 31


The data processing system is an important auditing tool used in verifying and recording each step in the toll collection process. Therefore, the computer software and hardware utilized should be flexible enough to generate data records and audit reports to meet the specific needs of an individual toll system. Shown in the third section of the diagram are the resulting data records that are generated after each step of the collection process. These records give the internal and external auditors of the A VI toll system the transaction details necessary to complete the audit reports. Points in the collection process where individualized audit reports are prepared are noted in the fourth section. Second, in the automatic lane collection process (Figure 11), the only difference from a manual process money flow is that the toll money is deposited by toll patrons into vaults contained within the automatic coin machine (ACM). The vaults are then transported to the money room by a bonded carrier. The bonded carrier may be a courier service or a toll supervisor/official. Also, within the data processing section, the coin machines are linked by computer to the lane controller. In this collection process, the coin vault construction and safety features become important auditing controls. Finally-, the third collection process (Figure 12) is the A VI (ExpressTOW Jane collection process. In the money flow section of this process, the collection and deposit of toll revenues are handled electronically. The handling of hard currency is eliminated through the electronic transfer of funds between the patron's bank and the toll operator's bank. This electronic process helps to reduce the loss of toll revenue for A VI transactions by eliminating the need for cash drawers and coin machines in express A VI lanes. It also eliminates the need to count, sort, and deposit monies collected on these lanes. 32


FIGURE 10 Manual Lane Collection Process ...... . ...... ., .,_ ., ,..1'0.. t.b'l ... Room .,_ e.c.. ...... ...... ..... .... .... o ...... .,....._., .... ,,..... """"'"' ....., '"". -.. --w...... eo.... .. [I](!) = h-llB' ----c....-.. m . ...::::... .... ,.II !.....! -<-(-lt.,O$to<-(.;(Ao(IOI'I 51.-0.,0111 t{,-,,.. ....... :::.: ..... I ::::: -._, SOURCE: Properly ol 1!-470 Public Highway Aulhorily """""""' ..... ._, ........... ......... -.. . . .. .... ... ... . ... .. .. .. __ .. ..... c. .... PHYSICAL FLOW OF MONEY I ....................... ._ : ... u., .. 1ltn : ............. ......... . . (.;;;;i') DATA PROCESSING """' o-tit """'' DATA RECORDS AUDIT REPORTS


--. ......... ......, ...... c... c....s. v ....... ---era. <3=-flow 1) .... ... --"'IOj?--1...1 __ -. ..... ... ....... -FIGUR E 11 Automatic Lane Collec ti o n P rocess e> _, ... ---C..P'' ... ... -' m:-. -...... c.. .. SOURCE: P roperly of E 470 Publi c HiaJoway Aulbocily PHYSICAL FLOW OF MONEY ............................. ............ . _Jliii_......... : : .... .,.,... .. -.. --..... .. -.. DATA PROCESSING -DATA RECORDS AUDIT REPORTS


Eap .. JOLL l 1MMdlc:nl "'4' -r CD IJJ RMdoefiCOI*ahf c c::D=f .... ll(\! l.w -......--ll[:::J-_,. ..... FIGURE 12 AVI (EspressTOLL) Lane Co ll ecti o n Process PHYSICAL FLOW OF M O NEY .....--.-----._, .. -;2 I g .. .. ,. 8AN( - j_-------------. ..... ceo ----DATA PROCESSING ---DATA RECORDS AUDIT REPORT S S OURCE: Proper t y o f E-470 Public Hi&hwa y Authority


Patron Account lnfoonat!on As an A VI vehicle enters a toll plaza lane, the A VI system detects the approaching vehicle and prepares for the transaction. When the vehicle passes over the lane tread l e, the system registers the vehicle by reading and identifying the assigned A VI code. As the code is read, the system verifies that the patron's account balance is sufficient to cover the toll charge. At this time, any lost tags, stolen tags, or below-balance accounts are registered. If the A VI code is associated with a valid account, the lane's traffic signal turns from red to green If not, the light remains red, indicating either a misread, a lane violation, or a problem with the patron's account. This necessitates the manual payment of the toll by the patron either in the automatic coin machine or to a collector. Once the vehicle is cleared to leave th e toll plaza, the system automatically updates the patron data base files and other associated files. All lane transactions are detected and recorded, whether the Jane is in an operational mode or closed. The processing of a standard A VI vehicle take s seconds to complete. Reports can be created to detail patron account information and patron account transactions. For example, Figure 13 is a flowchart from Amtech detailing the daily cash . reconciliation at the TollTagN Office (TollTagN is a registered trademark of Amtech). The office is the equivalent of an A VI Service Center. The Transaction by Type and Time Reports detail all daily entries by each employee's cash drawer. The individual Cash Drawer Reports reconcile each cash drawer's transactions to the rela ted types of deposits. The Office Daily Report is a total summary of all the Cash Drawer Reports. Another example of the type of A VI system reports that can be generated is illustrated in Figure 14 entitled Nightly System Reports. These reports can be run by the system at the close of business each day. Nightly System Reports reconcile and keep a historical record of all electronic transactions, charges to patron accounts, and changes in patron account balances. Ad Hoc Reports (Figure 15) list reports detailing account histories, individual tag information, account activities, and license plate information. The last example is Special Action Reports, Figure 16. These reports can be generated to show patron accounts with fund balances below minimum. Other reports can detail tags that have expired or patron credit cards that are rejected or expired. 36


AJI daily entries printed b y cash draw Reconciles 41acl'l cash drawer's transac::lons to credit card/cash it Grand sumnwy at all cash drawer FIGURE 13 T ransa::cn by TcatU:IIIln by T)1l8 RA!pllt rlll18 Repcn I _I r Alpoft C._ 2 Rtpoll + caaam.1 fllpoit + TaiTaq"'Oib oaiyRip:M 37 Summaries by T.ansac:ion Type security deposits rafunds Summaries by Payment Type ad card cash bil-to accounts Gtand Sc.unmaty at Orasu ars l:ly li31 ISa:!ion type l:ly payment type


Historical r8Q)I'd of all elearonic coHadions Reconciles totals tor each lane Shows nightly charges to pre-authoriZed aa:ctJms Used to post all totals for each day Used to trac:X all cnanges in balallc:U for the endre montn FIGURE 14 Daily Tol I !sage Aepcn Chaiji n Balli a Report 38 Sorted by: l ane shiltltime Shows: name phone number card type and l'1l.llrlbW expiration date, etc. Shows: beginning balances al c:hangee Shows: beginning balanca all changes encfng balanca


Used to document a single ac:caunt history Used for historical purposes (voluminous report) Used for historical purposes (voluminous report) Shows all monttHo-date activity by transaclon type tor-each account (voluminous report) license plates to TolfTag"' FIGURE lS 39 Shows each account transaction sorted by time Shows all account information Shows all related Toi!Tagno into""ation Shows summarized accotJnt of all transac tlonal data Shows: license plate ToUTag"" number


Used to in!onn "low Salanca" Cash pauons of ao:ount stall IS Used to contact patn)ns with "'fme ToiiTags"' Used to inform palrCns of ctedit limit or other auU'Iorizadon failures Used to obtain new expa l'lllion dales tor csldit c:vd:s FIGURE 16 Low Balanca Repon (Daily) Shows: name phone numbers CJmM1t balance ell:. Shows: name phone numbers expiration date etc. Shows: name phone numbels Cl edit card type ara:l numcer rajec:ien reason etc. Shows: name phone numbers credit card type and number date etc.


Manaeement and Operational Reports The reports generated by an A VI toll system can vary depending on the toll operator's accounting and information requirements. The following is a list of reports that can be compiled through information contained in the A VI computer system data base: individual lane, plaza, and system-wide transaction and status reports individual lane, plaza, and system-wide accounting and audit reports of toll r evenues system-wide electronic mail networ k maintenance and diagnostics reports for system equipment traffic statistics and traffic management reports daily and monthly system accounting reports daily and monthly A VI Service Center reports daily and monthly personnel status and payroll reports daily and monthly s ystem audit reports of Toll Assets Revenue collection systems are inherently subject to abuse; therefore, measures to safeguard revenues should be instituted. Within an A VI system there are seve ral key safeguard features. One is the ability to have video surveillance cameras in place to record lane violations and possible collector violations The computer hardware and software networking characteristic of video cameras is a substantial safeguard. System gene rated equipment status reports and diagnostic and maintenance reports are other safeguard features. Also, characteristic of the appropriate computer system are sign-on access procedures for all levels of personnel. The employees can gain access only to pre determined system levels assigned to their specific job requirements. Even toll collectors need to sign-on for shifts us ing their individualized identification codes and passwords. Most important is the ability of the system to collect tolls through t he electronic t ransfer of funds which elimi nates the extra precautions and costs needed to safeguard bard currency and re l ated data records Unrelated to the implementation of A Vl, t he Florida Turnpike bas already committed to the purchase and installation of an electronic accounting system and s urveillance equipment with current bond revenues. 41


Conclusions Several conclusions can be made with respect to operational and accounting functions. Toll revenue tracking for accounting and auditing purposes is enhanced by electronic fund transfers. The A VI system accounting hardware and software increase the accountant's auditing capabilities and sources of reliable system information. The electronic accounting and surveillance system to be purchased by the Florida Turnpike would need to be evaluated for A VI compatibility. Payment Systems In evaluating payment systems, there are three areas to be reviewed: toll structures, pre payment vs. post-payment, and actual methods of payment (cash, check or credit card). First, the Florida Turnpike's present payment system is summarized. Next, each of the three areas above are discussed according to ease of implem entation, expe riences of exist i ng A VI systems, and patron and Turnpike perceptions. Presently, the Florida Turnpike system does not utilize premium or discounted tolls or toll payment using major credit cards. Patrons can pay tolls with cash, traveler's checks, or personal checks, with personal checks accepted only as a last resort. Also, current policy allows only commercial traffic to prepay for future trips through toll p lazas by means of a prepaid account card. A non-revenue card is issued to users exempt from paying tolls. Post payment is accepted only when a patron does not have the proper fare available when passing t hrough the toll plaza (an Insufficient Funds Report is completed by the toll collector and a copy is issued to the patron for submission at a future date) or when a patron submits tolls resulting from previous lane violations. Pre-payment vs. Post-paymentAn A VI system requires either pre-or post-payment of tolls. The operational procedures for these two methods are quite different. Within an A VI pre-42


payment toll system, A VI users establish individual accounts with a prepaid balance either by cash, check, credit card or electronic funds transfer. Subsequently, when the A VI user passes through a toll lane with A VI equipment, the toll amount generated is debited (subtracted) from the user's prepaid account balance. The post-payment system operates differently It is based upon a billing process whereby the A VI user is charged based upon actual usage of the toll system in the preceding month. In evaluating pre-payment and post-payment toll systems, there are numerous implementation costs and issues to be considered. A pre-payment system will require the establishment of locations for opening and replenishing A VI accounts. Locations can be at various fixed office sites or advenised mobile van sites. Accounts opened and replenished through the mail can also be processed at the office sites. At a fiXed office site, an A VI Service Center could be established. Costs associated with the center would include but not be limited to construction costs or leasing costs, operational costs, computer costs and maintenance costs Another opt i on is a mail program where opening of accounts and distribution of tags could be accomplished through the mail. The mail program would still require a central processing location. A final option is mobile registration where a mobile unit could be dispatched to register patrons. When considering a post-payment plan, additional operational costs would need to be evaluated. The additional costs include monthly statement account mailings and collections on delinquent accounts receivable. The collection problems and costs associated with delinquent accounts far outweigh the benefit of correcting errors before statement billings are mailed. The collection process is viewed by many to be time consuming and undesirable. In reviewing existing A VI toll operations as outlined in Technical Memorandum 1, the San Diego Coronado Bridge is one of the few systems to offer a post-payment option. They were the only reviewed A VI operator offering this option and they have not indicated any major collection problems. However, the San Diego system is a demonstration project only and bas never been implemented on a full scale. Upon review of patron perceptions, most potential A VI users prefer a pre payment. This is evidenced by a large percentage of Turnpike users rejecting the post-payment option in 43


the survey research. They cited collection problems as being the main reason for this preference. Also, current patrons did not want the Turnpike to become a collection agency. Pa.pnent Methods A VI-generated tolls can be paid by cash, check, electronic funds transfer, or credit card. In addition, tolls can be either pre-paid or post-paid (billed). Because of the problems associated with a post-payment system, evaluation of th is option will not be considered any further. Pre-payment of A VI-generated tolls by cash requires that the patron visit an A VI center once to open an account and subsequently to replenish the account. Pre-payment by check can be accomplished either by visiting the A VI center or by using a mail program. Pre -paym ent by electronic funds transfer requires that the patron visit an A VI center only once, to open an account. An agreement is signed that allows funds from the patron's bank to be automatically transferred to his A VI account to replenish it when a pre-set minimum balance is reached. Pre-payment by major credit card is the same as by electronic funds transfer, except the patron's account is replenished by charging his credit card. Since the use of electronic funds transfer and credit cards does not require the handling of money, operational costs are reduced and implementation and maintenance of the payment program are simplified. Currently, all operators of existing A VI toll facilities accept cash or check for payment of tolls. These are the most popular toll payment options. The number of existing toll facilities using credit card or electronic funds transfer varies. Furthermore, Amtech, the operator of the Dallas North Tollway, offers their patrons the option of opening an anonymous account where all transactions can take p lace without providing the A VI toll facility with personal information. Finally, patron preferences would need to be studied before implementing any A VI system. The Florida T urnpike patrons surveyed chose credit cards as the least preferred payment option, primarily due to concerns about the confidentiality of account information. The perceived problem is that a government agency cannot properly handle and secure account 44


information. The most popular payment options were by cash or check tbrougb tbe mail. Conducting business by mail was popular because respondents perceived traveling to an A VI outlet as being inconvenient Toll Structures There are three A VI toll structure options. The first option is to charge premiums in addition to existing tolls. Advocates of premium tolls believe users of A VI should pay an extra charge for these special services. The second option is to discou n t existing tolls Advocates of discounted tolls believe users need to be encouraged to use A VI by offering discounts on services. They also contend that increased patron participation resulting from discounts would fully offset any declines in revenue. The third option is to keep the toll structure the same, neither offering discounts nor charging premiums Advocates of this final option contend that implementation problems would not exist and any governing legal documents would not need to be reviewed for compliance. In comparing premium and discount toll structures for ease of implementation and associated costs, they were found to be the same. An extensive marketing campaign would be needed to educa te potential A VI patrons on the new toll struc ture An account i ng and billing program woul d need to be instituted to keep track of patron toll charges and payments. Additional s ignage and literature detailing the new toll structure would need to be purchased, installed, and published Also, the Turnpike legal counsel would need to review legal statutes and any outstanding Turnp i ke bond issues to verify compliance. Furthermore even if t he premium or discount toll structures were not chosen, an efficient Turnpike A VI system should still consider all the above implementation issues and costs. Information is not widely available for evaluating toll structures based upon experiences from other existing A VI toll systems. As noted in Technical Memorandum 1, the Dallas North Tollway operated by Amtech is the only large-scale functioning A VI system charging premiums. The system seems to be a success. However, careful consideration must be given to area demographics, traffic volume, and unique system characteristics before applying this system elsewhere. On the other hand, the majority of other functioning A VI toll systems, la rge and small, offer discounted tolls. Most systems confirmed that discounts encourage A VI use. 45


In the final evaluation of toll structures, toll patron and Turnpike management attitudes and perceptions must be considered. Participants responding to the Florida Turnpike Patron Survey conducted during this study (Section V) fully supported the idea of discounts and were ambivalent to keeping tolls the same. This was evidenced by a large percentage of both mail-in and focus group survey respondents willing to use A VI if discounted tolls were offered or tolls were kept the same. However, the percentage willing to pay if premiums were charged was quite smalL Based on the survey results, the Florida Turnpike management has rejected the option of charging premiums for A VI usage, but they have not eliminated the consideration of offering discounted tolls in addition to keeping the present toll structure. In addition to the consideration of survey results, it is useful to consider historical data concerning transportation demand elasticities. Typically, the price elasticity of demand associated with the demand for transportation is relatively inelastic. An inelastic demand indicates that a one percent change in price will result in a less than one percent change in the quantity demanded. Therefore, despite the positive reaction to discounted tolls in the survey research, it is expected that a minor toll discount would not significantly increase participation in the A VI system. However, the use of discount tolls could still play an important role in a successful marketing program if deemed appropriate by Turnpike officials. Qther Issues -As with all A VI systems, whatever the payment options offered, the proper installation of the tag (decal or transponder) in the vehicle is essential. Instructions should be given or sent to the user when an A VI tag is purchased. In addition, an A VI system test lane should be set up at the A VI Center that allows the patron to test tag placement before using the tag at an actual A VI toll plaza. Some A VI operators insist that trained personnel install the A VI tag for the user, thereby reducing the probability of misreads due to faulty tag placement. Another issue to be considered is the establishment of a 1-(800) hot line phone number so that trained personnel could address problems and field questions from current and potential A VI users. Also, the toll agency should consider the possibility of establishing a direct computer link berween the Department of Motor Vehicles and the A VI system. This 46


would allow new and current residents the option of signing up for A VI when they register their vehicles Finally, if choosing to accept payment by credit card, negotiations would need to be made with the banks that conduct business with Florida's Turnpike to detennine r easonable transaction fees. Conclusions Respondents of the Florida Turnpike Patron Survey prefer pre-paying tolls. The Thrnpike should encourage credit card and electronic funds transfer as toll payment options; however, the Thrnpike needs to assure potential A VI users that the privacy of their accounts would be safeguarded. These options require less operational costs and result in improved accountability and convenience Post-payment of tolls should not be a payment option since it leads to collection problems on delinquent accounts. Premiums on tolls should not be considered since survey respondents were strongly against them. Trained A VI personnel should install the A VI tag (transponder or decal) to reduce resulting misreads at the toll plazas. LEGAL ISSUES Florida's Turnpike operates under Florida Statute Chapter 338, "Limited Access and Toll Facilities," which gives the authority to establish, regulate, and operate limited access facilities. When considering the i mplementation of A VI, legal limitations contained within 47


outstanding bond indenture agreements and present traffic infraction statutes need to be addressed. Also, to support the prosecution of toll violators, A VI equipment specifications need to be reviewed. The legal issues and bond indenture limitations discussed below are not represented to be aU-inclusive or held out to be definitive legal opinion. Payment Options Paying for A VI service either by credit card, electronic funds transfer, check, or cash does not pose any legal problems. The options of pre-paying tolls or paying premiums in addition to existing tolls do not pose any legal problems either The premiums discussed here are not contingent upon distance travelled Only when evaluating the options of post-payment and discounts does the le gality of these methods need to be considered. The option of post payment does not conform to the legal directive given to the Turnpike of collecting all toll revenues it is entitled to receive. However, support expressed by patrons surveyed regarding the discounting of tolls suggests that further investigation is required concerning relevant legal issues. Discounts are mentioned in the legal wording of the outstanding Florida Turnpike Bond Indenture agreements. In the body of the agreements it states that the Department of Transportation is without independent power to reduce toll rates or remove tolls from all or a portion of the Turnpike System, except when certain conditions are met. This statement is legally binding until all the outstanding bonds and associated interest are fully paid and discharged. This does preclude the conversion of one system of toll collection to another system of toll collection. In the instance when tolls are increased, reduced, or removed, the action must be based on a survey and certified recommendation of t r affic engineers For example, the current plan is to raise the existing toll rates by 50 percent over the next five years in order to meet needed revenue l evels. Recommendations from the traffic engineers would certify that the reduction or removal of tolls would not affect each fiscal year's funding requirements summarized in the foll owing equations: 48


Gross Revenues (each year)(100% x Cost of Mainteaance and Operations) = Net Revenues Net Revenues (120% x Annual Debt Service Requirement of Bonds) (100% x Other Expense Requirements of the Bond) 4 0 Where: Gross Revenues = "All tolls, revenues, rates, fees, charges, receipts, rents and other income derived from and in connection with the operation of the Turnpike plus (unless negated) income from investments of funds and accounts aeated by the bond issuance plus proceeds of any usc and occupancy insurance." CO! of Maintenance = "All usual and ordinary coots and expenses which are the obligation of the Department to keep the Turnpike System open to public travel (excludes operational costs and non toll roads except feeder roads)." Cost of 01H'rations "All costs and expenses incurred from portions of the Turnpike System being a toll facility, includes costs of collecting and accounting for toUs, insurance, employee bood premiums, coosulting engineers fees, and other costs incurred from the system being a toll road." Debt Service Reserve Acoounl requires for each bood, the lesser of the following; 1) 125% of the average Annual Debt Service Requirement for curreot and succeeding ftseal years. 2) maximum anoual debt service. 3) 10% of the aggregate of the original proceeds received from the initial sale of all outstanding boods. 4) maximum debt service reserve allowed by the IRS for tax-exempl obligations. On or bef o re February 21 of each year, the Department of Transportation uses the above equations to determine if funding requirements will be met for the next fiscal year. If not, traffic engineers will conduct a study and produce a schedule of t o lls which will provide revenues sufficient to meet required fund ing and to replace any deficit The Department may i ncrease toll rates upon written recommendation of the traffic engineers. The Department can make other adjustments (including reductions) or rec l assifications of rates o r establish rates provided the rates meet requirements set forth in the Bond Indenture agreements and summarized as the following: 49


Enforcement 1) The adjustments or reclassifications are recommended and certified by the traffic engineers and affects traffic of a character specified by such engineers accounting for less than 10 percent of revenues. 2) The adjustments or reclassifications will not result in a reduction of Net Revenues for current or future fiscal years and the calculations supporting these numbers (estimates of revenues, costs of operations, and maintenance) are certified by the traffic engineers. A VI Jane violations are an important issue since time savings and convenience are the major benefits of A VI. A successful A VI toll system will need to utilize a high speed video camera system to deter A VI lane violators. Most states where A VI systems are located have passed statutory laws enabling A VI violators to be legally identified and cited based on videotape evidence. Colorado is an example where specific legislation was passed for the prosecution of A VI violators who will be using the new E-470 toll highway. Also, the City of Pasadena, California, is currently enforcing speeding tickets by using a photographic system called "PHOTO-COP." This system photographs speeders and sends a pre-printed citation through the mail. However, state laws allowing pictures generated from remote video cameras to support prosecutions are not uniform. For example, A VI patrons incurring violations on the Dallas North Tollway were sent numerous letters detailing their violat ions It took Dallas patrons only a short time to realize the legal evidence generated from the Tollway's video cameras was not supportable in the state court system. The statutory laws i n New York are another example. New York statutes detail the need for an "enforcem ent agent" to witness the patron avoiding the payment of toll and to issue a summons at that time. Florida has a similar situation. The infraction must be witnessed by an office r and the citation must be i ssued to the driver of the vehicle avoiding the toll and not the registered owner of the vehicle. so


Florida Statute 316.1001 details who must pay tolls and the penalties required for toll avoidance. The statue reads: Payment of toll on toll facilities required; penalties.-No person shall use any toll facility without payment of tolls, except as provided in s. 338.155. Failure to pay a prescribed toll is a noncriminal traffic infraction, punishable as a moving violation pursuant to s. 318.18." Note that s. 338.155 details persons who are exempt from toll payment and s. 318.18 details the amount of civil penalties levied for not paying tolls Any person cited for this type of infraction must sign and accept a citation from a witnessing offic er. A change in state law would need to be proposed to the Florida State Legislature to allow the admissibility of evidence collected by videotape from A VI video cameras. Also, legislation is needed to charge the vehicle's registered owner with the penalty of avoiding tolls. This eliminates the need for the video camera to specifically identify the driver of the vehicle Video Camera Eauioment Video camera equipment is an important part of the A VI enforcement system and specifications should be reviewed carefully. All lanes should contain remote control, high speed video cameras. This enables the recording of all lane violations, whether the lane is open for operation or not. As a violation occurs, the camera is activated. It records the offending vehicle, its license plate, the lane's traffic signal, and the violation indicator as the vehicle travels over the lane's exit loop Cameras in operation now can provide legible vehicle license plate pictures for speeds of up to 60 miles per hour. The specific capabilities of the video cameras and related equipment should be flexible to meet the needs of various A VI systems. However all video enforcement systems shou ld detail on the video picture the date, time, and lane number for when and where the violation takes place. The system-generated video picture should also be rec orded on a computer disk and retrieved when a hard copy of the picture needs to be reviewed. The software that drives the operation of the video lane cameras should trigger the cameras to record not only when non-payment violations occur but also when a vehicl e passes that is 51


using a reported lost or stolen tag, when the patron's account is below a predetermined minimum balance, when the account bas been suspended, or when the axles classified do not match the treadle count indicated. It is suggested that the recorded videotapes be reviewed by FDOT personnel before violation notices are sent to patrons. The review process will allow any misreads to be noted and corrected. Some systems record the number of violations each patron has incurred and sends notices to repeat offenders only. It is also suggested that a predetermined level of acceptable violations would need to be set by FDOT. In addition, procedures would need to be determined when the video picture of the license plate is not readable. The FDOT should investigate the possibility of gaining access to the Florida Department of Motor Vehicle's data base via the Florida Highway Patrol. This would enable FOOT personnel to match each violation with the description of the offending vehicle and license plate A few violators will never respond to a letter detailing their infractions unless held to by a court of law. General consensus among present operators of A VI enforcement systems is that the average patron pays any outstanding tolls when presented with a photograph of their vehicle incurring the violations Conclusions Present Turnpike Bond Indenture agreements discourage the use of discounts and need to be reviewed further for legal compliance if discounted tolls are to be offered. Legislation will need to be introduced to the Florida State Legislature recommending that toll violation penalties be charged to the owner of the vehicle as opposed to the driver. The video camera enforcement system installed by the Turnpike will need to meet rigid specifications that optimize features of proper lighting, security, software and hardware integration, weatherproofing, and read reliability. 52


Predetermined levels of acceptable patron violations will need to be determined The Florida Turnpike will need to further study the option of interfacing the A VI enforcement system with t he Florida D epartment of Motor Vehicles and Florida Highway Patrol. COMPUTER SYSTEM REQUIREMENTS Due to the unavailability of computer description data, the existing Turnpike hardware and software components could not be evaluated. Also due to confidentiality at the time of the preparation of this report, the proposed Turnpike computer system upgrade specification was not available for review. However, this section provides an overview of the basic computer system requirements for the accommodation of A VI. The design and configuration presented indicate that A VI can be added as an upgrade to a well-designed, modem toll plaza. A "well designed modem toll plaza" is defined as a toll facility with state-of-th e -art microprocessor-based manual and automatic coin machines, a microprocessor-based lane controller that is programmable, a plaza computer system, and a host computer system with network capabilities. These features represent the latest designs being implemented today at new or retrofitted toll facilities. The Digital Equipment Corporation (DEC) VAX computers are the most common models f ound in the A VI/toll collection industry. This is due to their power, size, networking capabilities, a nd architecture. They also provide a consistent operating system and programming environment to protect the software investment of toll agencies. Additionally, DEC has developed computer architecture that provides for hardware growth and expandability. For example, the DEC VAX 6XXX family of processors provides for symmetrical multi-processing (SMP) of one to six internal processors. The unit can be expanded by the simple addition of a second to sixth CPU board for 8 to 64 units of processing power. This is an important factor in adding office automation, Computer Aided Design, and other traffic management software to the host computer. With other computer systems, such expansion requires the entire replacement of computer systems, added facility 53


loads on power and air conditioning, and expensive changes in operating system software. The VAX 6XXX family, the VAX 4XXX family, and the Micro VAX 3XXX line of computers provide long term protect ion and ease of upgrade. Furthermore, the common operating system, the VAX VMS Operating System, spans the entire line of processors for a common interface, programmer knowledge, and ease of interfacing. The dual processor configuration provides for high availability. While a high operating percentage of time i s valued with any one processor, the dual configuration provides for typically 99.6 percent or better uptime performance. Maintenance, system hardware upgrades, and system software upgrades can be performed on one processor while the other is functioning. After the upgrade i s test ed, the system can be switched over and the second processor can be serviced. Failures or problems will also cause an automatic switchover from one processor to the other. Downtime is minimized and can be calculated in minutes rather than hours. For the seven day a week, twenty-four hour a day, non-stop e nvironment of a transaction processing system for a toll road, this configuration i s a necessity The configuration for dual processors also provides dual power supplies, dual memory, dual network links and dual system disks. This redundancy eliminates many potential failures due to electrical or mechanical failures of components. At the host level, for a l arge size host configuration, the newer VAX 4XX:X series provides a good platform. At small toll facilities, the Micro VAX 3XXX line would match the requirements. At the pla2a level, the Micro VAX 3XXX series of DEC processors is adequate. The size and exact specification can be fine-tuned to the size and traffic volumes of the pla2a. Due to the sensitivity of the transaction data, the toll transactions should be duplicated on a second drive that is synchronized to the first data disk. This disk duplexing or "shadow recording" is common in transaction processing environments and protects system software. However, some hardware solutions are also available. The disk drives must be sized to accommodate the needs of the system. At the host location, the disks should be sized to accommodate three to six months' worth of data. Storage arrays today are 4 to 12 GBs of data, each disk providing in excess of 1 GB of storage. At a plaza location, 31 days of on-line storage is normal. Transaction data and files are again "shadow recorded" for protection. Data must be synchronized with the host 54


data to ensure data recovery in case of a malfunction. Disks at this level are typically smaller but can also range to the 1 GB size for large plazas. Data stored at this level include the transaction data files, the maintenance records of the plaza, the event logs, security files, and the A VI account li st. At the lane controller, adequate storage is necessary for two or three days. This storage must also be available to handle the transaction files, the maintenance data files, and the A VI account verification files. One of the greatest shortcomings in the industry today is inadequate st orage space provided in the lane controllers. The computer system requires an extensive network to constantly update file data and pass transaction, maintenance, diagnostic, and electronic messages from one processing l evel to the next. The speed of the network and the ability to access information at all levels is essential to a toll facility. DECnet or Ethernet are good networks for the local and wide area needs of the toll authority. Network interfaces will range from the local area network for attached devices or lane controllers to wide area networks that tie together all the plazas in a given system. All the local devices attached or serviced by the host computer are interfaced through a local area network (LAN). The co mmon network media is Ethernet for DECnet or TCP /IP protocol. Routers and servers today will service both protocols on the same thin or thick wire installation. The dispersed plazas and work stations are serviced by the same protocol over a wide area network (W ANS) This typically requires a network router and a high speed data circuit. Digital data circuits are becoming more practical and affordable for W ANS. Fiber Optic installation provides the greatest flexibility, and newer facili t ies should look to the employment of their own fiber optic bac kbone rather than leased circuits for wide area coverage. Long-term backup can be accomplished with the new 4mm DAT cartridge tape recorders. These 4mrn DAT tapes are superior to the old reel-to-reel tape drives in size, cost, and storage capability. The additional advantage is the physical size of the cartridges which are smaller than the common stereo cassette cartridge. This size allows more storage i n a smaller physical space than the older reel-to-reel tape racks which took entire rooms to store. The selected backup media should be consistent with the backup media placed on the p laza computer system. Autoloaders and stackers are now available for 4mm OAT tape 55


systems for loading tapes in series or in an unattended backup mode. Due to the controllers of the 4mm OAT tape drives being slaved to a single CPU, two 4mm OAT drives, one for each processor, will be needed. One could be eliminated if a 4mm OAT tape drive can be interfaced to both processors through a hardware interface or be connected on the network. High speed printing can be accomplished by either impact or laser printers The use of laser printers is growing and becoming commonplace at toll facilities. Tbis is based on the cost, speed and quality of the current laser printer technology. The toll facility can expect to produce 30,000 to 50,000 pages of printed material per month This heavy volume dictates a high speed impact printer ( 600 to 2,000 LPM) or a laser printer that is 20 to 40 pages per minute. Typically, a high speed impact printer and a laser printer will be utilized for the printing needs of the facilities. With rational databases to retrieve information, most data inquiries can be produced on CRT screens; however, reports billings, and maintenance logs will generate printed material. Adequate printing services must be provided in the system configuration. PC workstations, engineering workstations and terminals are all interfaced through the LAN. Personal computers and workstations will normally require a server or network interface unit (NIU) for connection. Terminals and printers will require a different type of server designed for terminal serving. Newer X-Window terminals may have the Ethernet thick or thin wire connector internal to the unit and thereby negate the need for the NIU. The number and types of terminals, PCs, and workstations will depend upon the needs of the toll agency. Interfaces to other agencies may require a gateway device to talk to different types of computers. In a heterogeneous network, gateways are installed to translate between computer systems The simplest interface is typically a 1500 bps tape that can be exchanged between computers. Tbis will dictate a reel-to-reel tape unit to be added to the hardware configuration. With the above understanding, the addition of A VI to such facilities is a simple and cost effective step. The three-tier computer network (host-plaza-lane), will carry the A VI transaction load The transaction processing will require only minor modification to encompass the additional A VI transactions. For a fully-integrated system, the addition of 56


A VI should simply add another means o f collecting fares from patrons A VI can be visualized as the replacement of existing charge accounts without the manual transaction of magnetic strip cards commuter tickets, or visual stickers At the lane level, the only additions necessary are the A VI reader /antenna units, reader I controller unit, cabling, and software interface to the Jane controller Since these items should be low power consumption units, excess uninterruptible power supply (UPS) should handle the new loads. If the A VI system to be installed is a high power A VI system, then the system integration will be compounded by power and shielding requirements The addition of A VI to an existing, well-designed, modem toll plaza will also impact the three-tier computer network. These impacts are the network loading, the processor memory r e quirements, the on-line storage, and the off-line storage of the processors located at each of the three levels of the computer system. The network loading will need to be studied prior to installation Higher throughput rates will be seen and this will impact the network response times I n a manual or automatic coin lane where a transaction can occur every four to ten seconds respectively, network massaging has ample time to transmit and receive digitiz e d packets of information A VI on the other hand could produce a 1 5 second per transaction speed, and will require higher communication speeds. With the addition of A VI will come the account l ookup and verification of accounting information. The impact of this processing load will need to be assessed on the lane controller, plaza computers, and host computers The greatest impact may occur at the lane controller level where real time processing of the transaction may be adversely affected due to a lack of CPU processing. At the h ost level, the accounting and billing will impact the CPU l oad of the processor s in ce the billing and accounting for the A VI transactions will add additional tasks. These could be h andled on a backup processor while the primary processor is handling normal functions. In the case of a processor being down, the A VI processing may affect the response time of the host system when running A VI and normal toll processing on the same computer. If the i nit ial design accounts for ample growth and excess capacity, then little to no additional CPU costs or memory will be necessary. 57


The on-line and off-line memory requirements will be impacted by the addit ion of A VI The A VI account data, billing data, and transaction information will require additional memory for all three levels of the computer SYStem. The lane controller is the least impacted. Its storage of transaction information will remain the same Only the additional A VI transaction information will represent a percentage of the total number of transactions. In some real-time systems, the A VI verification file is kept at the lane controller. This allows stand-alone processing when the communications or p l aza computer are down. In this type of installation, lane controller memory will need to be expanded to handle the size of the A VI account list. The plaza computer storage will be impacted similar to the lane controller The host computer's on-line and off-line memory requirement will be impacted the greatest. The host computer will require A VI account billing and long term storage of account information for billing purposes. This storage will require additional disk space The configuration for the host and plaza computers specified dual or shadow recording of on-line transaction data. Therefore, two additional disks and controllers could be necessary Size of disks will be dependent upon the size of the facility. Disks with 1 to 2 GBs of data storage are becoming commonplace and affordable. Off-line storage will also need to be expanded at the host site. The specified 4mm DAT cartridge tape drives should prove ample storage with their 1.2GB capabilities (Note: new data compression techniques are being combined with the 4mm DAT tape drives and 2 to 2.5 GB of storage are becoming available on the same tape used today for 1.2 GBs of storage). The number and storage requirements of the tape cartridges will require ample storage space in the planning of new facilities or additions to existing facilities. The air conditioning, power, UPS, and physical space requirements will be minimal and should be considered in the initial design of a new system. Recomm e ndatio n s During the next phase of A VI implementation, that is, concept design, it is recommended that a thorough assessment of the existing and proposed Turnpike computer system be initiated. Specific hardware and software requirements of A VI integration, as outlined above, can be clearly identified and included in the final design specification 58


OWNERSHIP/FINANCE OPTION S There are three major ownership arrangements to consider in the process of negotiating the implementation of an A VI system on Florida's Turnpike. This includes the agency owning and operating a system, a vendor owning and operating a system, and various lease agreements. E ach of these arrangements is characterized by numerous advantages and disadvantages. The purpose of this section is to identify these advantages and disadvantages for the Thrnpike management. Most A VI vendors are flexible with respect to the administrative arrangement that is selected for the ownership and operation of an A VI system However the ability of a toll agency to select an ownersbip/fmance arrangement is limited by the agency's charter. In some instances, an agency is not pertnitted to subcontract the responsibility of fare collection. This would limit the ability to subcontract A VI under the vendor operation scenario Al:!!ncy Owns and Operates In this ownership arrangement, the Turnpike would purchase the entire A V I system and then operate the system independent of the vendor All responsibilities wou l d be with the implementing toll facility with the exception of those responsibilities that would remain with the vendor as a result of warranties and maintenance agreements The advantages and disadvantages of this ownership arrangement are enumerated bel ow. The implement ing toll facility maintains complete control over the A VI system including operation, administration, toll collection, maintenance arrangements, and all other aspects of the system. Maintaining this type of control may be preferred by many toll facilities. The toll facility enjoys the interest earned on all pre-payments and deposits collected as part of the A VI program 59


J)isadvantaees The facility will be held fully accountable for the successes and/or failures of the A VI system since it maintains control over all aspects of the system. If the system does not succeed, the agency is locked in on the equipment already purchased and installed. Owning and operating an A VI system requires the toll facility to incur significant capital costs over and above the conventional toll collection system. Additional support staff is required to maintain and monitor the A VI system. Additional training of staff is required particularly as it relates to the production and analysis of the many information and audit reports that the computer system provides as a resulting of interfacing with the A VI system. Vendor Qwns and Ooerates This arrangement involves the toll authority hiring an independent contractor to administer the A VI system including the installation, operation, administration, and maintenance. A typical vendorownership arrangement would preclude the applicable toll facility from purchasing any A VI equipment or software. An arrangement could be made where a vendor could be paid a flat fee to install, administer, and operate the system or, once the system became fully operational, the toll facility could be required to pay a fee for each transaction processed through the A VI system. This fee could then be passed on to the patrons in the form of premium tolls. In addition, each A VI patron could be required to pay a monthly fe. e for the use of the A VI tag. This arrangement could also tie into a tease/purchase agreement after a time period agreed upon by both parties involved. This type of arrangement can be termed an operating or service lease since it'provides for financing, operation, and maintenance. One advantage of an operating lease is that it frequently contains a cancellation clause which gives the lessee the ability to cancel the lease and return the equipment before the expiration of the basic tease agreement. 60


Advantaees The toll facility would no longer be directly responsible for the successes and/or failures of the A VI system. However, the toll facility would be responsible for monitoring and evaluating the performance of the contractor. This necessitates a much smaller support staff for A VI within the toll facility. If for some reason the agency were not satisfied with the A VI system, it would not be as costly to remove the A VI system since the agency does not own the A VI equipment. The extent to which this is possible would depend upon the negotiated contract. There is some incentive to succeed if compensation to the independent contractor is directly tied to the level of participation in the A VI system The toll facility can define performance objectives/standards which must be met by the independent contractor. The typical independent contractor already has significant marketing resources to draw upon which increases the potential for maximizing patron participation. D isadyanwus The ability of the toll facility to collect the proper fares is dependent upon the private company's records of fares collected. Such a system would require detailed audit controls if the agency permits an independent contractor to electronically handle large sums of tolls collected as a result of A VI. Depending on the contractual arrangement, the toll facility may Jose the "float" on pre payments and deposits collected as a result of the A VI system Under this arrangement, the agency never owns anything. This could result in some problems over the life of the agreement. 61


Information concerning the A VI users and their billing records could be the property of the independent contractor and not the toll facility. Ownership of these records is not always clear at the end of the negotiated contract. The toll facility may suffer from an identity problem As an example, the A VI patrons on the North Dallas Expressway identify with the private vendor and not the toll agency. Lease A2ftl!lments There are numerous lease arrangements in which an agency could become involved. A lease arrangement results in sharing responsibilities between the agency and the independent contractor. The burden of these responsibilities would be negotiated in the lease contract. Probably the most popular type of lease arrangement gives the implementing agency the choice of purchasing the equipment at the end of the lease agreement Advanta&es There is a great deal of flexibility involved with a lease agreement The agency could be responsible for those administrative tasks that are believed to be appropriate for the agency while the remaining tasks could be built into the negotiated contract. The toll facility ultimately assumes ownership of the equipment under most lease arrangements. Disadvanta&es Under this arrangement, the agency does not own the equipment over the life of the lease agreement This can result in problems over the life of the agreement. 62


Historically, Florida's Turnpike has not had positive experiences with various lease arrangements. Rei:ommendations Flo ri da's Turnpike essentially has three options to consider for the ownership and financing of an A VI system: agency ownership and operation, vendor ownership and operation, and lease arrangement. The advantages and disadvantages of each option have been identified to assist Turnpike officials in making this policy decision. 63


V. SURVEY RESEARCH Survey research is a valuable tool for assessing and evaluating the perceptions, attitudes, and characteristics of selected market segments regarding new products and services. As a result, it was determined that survey research should be included in the analysis of A VI and its potential application on the Florida Turnpike. A VI is considered a service that has the potential to benefit the user, the non-user, the Turnpike, and the community as a whole. However, this potential will not be realized unless the Turnpike user market reacts positively to such a service. The surveys conducted by CUTR will help measure the potential of A VI on Florida's Turnpike. The use of survey research in this analysis includes the following five categories: surveys of existing A VI users, summary of other A VI-related surveys, Florida Turnpike patron interviews, mail-back surveys, and focus groups. Each of the survey categories are summarized and reviewed in this section. SURVEY OF EXISTING AVI USERS The original intent of this survey category was to establish a socio-

CliTR visited A VI system locations throughout the United States including Dallas North Tollway San Diego-Coronado Bridge, Crescent City Connection in New Orleans, and several systems in the state of Florida includ ing Pinellas Bayway, Treasure Island, Broad Causeway (Bay Harbor Island), Sanibel Causeway, and the Cape Coral Bridge. Although it was not possible for CUTR to conduct original survey research of existing A VI users it i s possible to summarize observations made as a result of visiting the A VI systems indicated above. In addition, the management of many A VI systems provided a profile of the typical A VI user based on their observations. It is important to note that the potential market for A VI use varies with each installation, given the socio-economic and travel characteristics of the area. However, some general observations can be made regarding the profile of existing A VI users. Most A VI users are in the middle to high income brackets and travel the facility frequently in their commute to work. The income factor indicates that these individuals are more likely to value their time more highly while the greater use of the facility indicates that these individuals will realize significant time savings. Also, if additional costs are required to participate in the A VI system, those in the middle to high income brackets are more li kely to pay the additional cost. OTIIER AVIRELATED SURVEYS Several A VI-related surveys were discovered in the process of gathering information for this project. In particular, three surveys were discovered that specifically survey users of toll facilities that do not yet have A VI. The intent of these surveys was to identify characteristics of current users of the facility as well as to ascertain perceptions and attitudes toward A VI. These surveys were conducted by the Virginia Department of Transportation, the 111inois State Toll H ighway Authority, and the Oklahoma Turnpike Authority. Vlrf:inia Department or Transportation As part of the Dulles Fastoll project, the Virginia Department of Transportation (VDOT) conducted a survey of existing users of the Dulles Toll Road to determine the market potential of A VI on that facility VDOT contracted with Castle Rock Consultants to 65


conduct the survey and analyze the results. Dulles Toll Road is located in a suburb of Washington D.C. The analysis was comprised of four distinct survey efforts including an in-depth interview, a survey of Flashpass system users (conventional decal system), a survey of potential commercial users, and a mail back/hand back survey of users of the facility. The initial survey consisted of in-depth interviews of 50 randomly selected nsers of the Dulles Toll Road. The objective of these interviews was to provide preliminary information to help in the creation of the mail-back/hand-back questionnaire. Forty interviews were conducted over the phone, while the remaining ten interviews were conducted face-to-face. The survey was designed to identify perceptions and attitudes towards the following: level of interest under five different A VI cost alternatives level of interest based on different traveling speeds that A VI would permit method of fitting the electronic tag t o a vehicle various payment methods detailed toll road usage statement The respondents interviewed were representative of the most frequent users of the facility as they traveled the facility approximately 11 times per week on average. Nearly 75 percent of the trips made by these respondents were trips to and from work. The remaining trips were typically fo r leisure or for other personal reasons. Very few trips were related to other business-related purposes. A statement was read to each respondent describing the proposed Fastoll system. Approximately 65 percent reacted positively, 25 percent had mixed feelings, and 10 percent responded negatively. Sixty percent indicated interest in participating in the system as described. Over 50 percent believed it would be important for the system to offer dedicated A VI lanes at the exit ramps while less than 33 percent believed this to be important at the mainline plazas. Thirty-eight percent indicated they were very interested or quite interested in bypassing queues at speeds from 10 to 15 miles per hour. When speed increases to 45 to 50 mile s per hour, interest increases to approximately 58 percent 66


The interviews also included questions regarding the impact of charging a fee for the tag. Sixty-six percent of respondents were very interested in a system with no deposit and no annual fee. This declines to 8 percent with a $50 deposit and a $10 ann ual fee. Approximately 60 percent of respondents indicated a preference to pay for their use of the system by check. Most of the remaining respondents preferred the nse of a credit card while a very small percentage preferred payment by cash. There was no strong feeling regarding a pre-payment system or a post -pa yment system; however, a post-payment system appeared to be slightly favored. Regular account statements were desired by 60 percent of the respondents. Second, an interview survey of Flash pass users was conducted to identify the perceptions of this specific group of toll road users. A Flashpass is a visual decal that patrons could purchase that would permit unlimited passage through Dulles Toll Road plazas for a period of a month. The Flashpass was eliminated after numerous problems were experienced. However, the Flash pass program was still in existence at the time this survey was conducted. For that reason, a separate survey of Flashpass users was included. The level of interest in the Fastoll project increased significantly from the random sample discussed previously. Ninety-seven percent of those interviewed believed it to be extremely important to have ded icated A VI lanes at the main plaza. Although the level of interest increased with speed, the level of interest was extremely high at all speeds. Ninety-four percent were very interested in using the system with a $50 deposit and no annual fee. Interest fell to 47 percent when a $10 annual fee was added. In general, the Flasbpass user survey confirmed the results of tbe in -dep th random sample. A third survey was conducted to determine the potential for commercial use of the Fastoll system on the Dulles Toll Road. To find regular commercial users, 129 firms were contacted in the area. Only 15 firms used the facility often enough to justify an interview The majority of firms indicated that they avoid the toll road for a number of reasons, including traffic delays and inconvenience These 15 firms had a total of 1 266 vehicles; 61 percent were automobiles and 13 percent were trucks. The remaining vehicles inc luded small pickups, vans, and buses. The fleet managers for each of these firms were the ind ividuals who responded to the questions in the i nterview. 67


In the final survey, 10,050 questionnaires were distributed at Dulles Toll Road plazas of which 2,538 were returned, resulting in a response rate of approximately 25 percent Thirty five percent indicated they would definitely panicipate in the system immediately even if it required a $10 deposit With a $10 incentive ($10 in free tolls), over 50 percent indicated they would immediately panicipate in the Fastoll system. The survey methodology also pertnitted the identification and consideration of various market segments and their potential for panicipation in the system. Respondents were given the opportunity to make additional, unsolicited comments regarding the Fastoll system. Of the 2,538 respondents, 940 took advantage of this opportunity. For the most part, these comments were questions that respondents had regarding the operation of the system, including concern for safety and consequences of lost, stolen, or damaged tag. There was also some concern about the possibility of the system misreading tags and charging improper accounts. Other comments indicated a preference for transponder mobility; that is, users with more than one car wanted the ability to transfer a transponder from one car to another. Tbe majority of negative comments stemmed from those users who believed they should not be paying tolls at all. Very few negative comments were received r egarding the actual Fastoll system. I n summary, the majority of comments were encouraging and supported the efforts of the Fastoll project. Illinois State Toll Hi&hway Authority In early 1989, the lllinois State Toll Highway Authority conducted a relatively short survey of its toll facility users to determine their level of interest in an A VI system; 1 119 successful surveys were completed. Over 69 percent of respondents indicated they were either very interested (36 percent) or moderately interested (33 percent) in an A VI system on the Illinois toll roads The remaining 30 percent were not interested in A VI. Respondents were also questioned regarding the amount they would be willing to pay for the transponder necessary to participate i n an A VI system. Approximately 50 percent were willing to pay between $20 and $35. This willingness to pay drops to 12 percent when the cost of the transponder is between $35 and $50. Only 4 percent are still willing purchase a transponder if the cost were between $50 and $65 68


Oklahoma Turnpike Authority The Oklahoma Turnpike Authority recently conducted a survey of turnpike patrons to determine their attitudes toward A VI and to determine some basic user characteristics. A short, five-question survey instrument was distributed to 15,000 vehicles on the Turner Turnpike and to 15,000 vehicles on the Rogers Turnpike. Of the 30,000 distributed questionnaires, 3,003 were returned, resulting in a response rate of 10 percent. The sli_.[Vey instrument was designed to establish how often patrons traveled on the turnpike, the purpose of the trips, and the level of interest in A VI. Based on the survey results provided, 56 percent of the 2,688 useable responses indicated that they were interested and would likely use A VI if it were provided at these locations. FLORIDA TURNPIKE PATRON SURVEYS The Florida Turnpike patrons surveys consist of three parts: in terviews mail back surveys, and focus groups. Interviews were conducted by CUTR at two mainline barriers (Tamiami Plaza and Golden Glades P laza) as well as at six exits (Sunrise, State Road 84, and Commercial Boulevard in Ft. Lauderdale and Kissimmee/St. Cloud, Orlando South, and Interstate 4 in the Orlando area). The mail back survey instrument was handed out at these same Turnpike locations. Finally, two focus groups were conducted by CUTR at a convenient locatio n in South Florida (Dade County). The selected sites for the interviews and mail back survey are illustrated in Figure 17. The methodology for the selection of these sites for survey research is described below. 69




Selection or Survey Locations The selection of survey locations was determined from traffic survey data contained in the April-May 1988 report prepared by URS/Coverdale & Colpitts. The primary objective was to select survey locations that would capture the greatest percentage of five (or more) day per week users, or the basic commuter on the Turnpike. According to this report, the top ten interchange entry/exit points for five (or more) day/week users are as indicated in Table 4 TABLE4 Top Ten Entry {Exit Points On Florida's Turnpike With Daily Number of Five Day Per Week Trips 1. Golden Glades 14,348 2. Commercial Boulevard u,m 3. Sunrise Boulevard 10,038 4. HoUywood 9,585 5. S.R. 84 (l-595) 9,217 6. Lake Worth 5,830 7. West Palm Beach 4,905 8. Pompano Beach 4,835 9. Interstate 4 4,664 10. Delray Beach 4,321 Just as important as distinguishing those points of entry/exit with the greatest perc entage of five (or more) day per week users, isolating those trip links with a high percentage (greater than 50 percent) of five (or more) day per week users is also necessary in selecting the fewest but most appropriate survey locations. Origin-destination data was obtained from the traffic survey report. Figure 18 illustrates the major entry/exit points and trip links for the commuter portion (five or more day/week users) of Florida's Turnpike u sers. 71


fjulf of BEE UNE WEST __.. EXPRESSWAY ORLANDO FUTURE POINCIANNA INTERCHANGE FIGURE 18 TURNPIKE COMMUTER TRAVEL PATTERNS G) TOP TEN ENTRY/EXIT POINTS FOR 5 (or more) DAY/WEEK TRIPS I TRIP LINKS WITH GREATER THAN 50%5 (or more) DAY/WEEK TRIPS NOTE: Since H.E F .T. Is not on ticket syslem, I raffle survey data could not be oblalned. J11tfantic Ocean 0-------((({,((, _ ------m-FORT LAUDERDALE MIAMI HOMESTEAD N 0 5 10 ... 20


Figure 18 indicates that the largest portion of Turnpike commuter traffic is located in Broward County, with smaller portions in Palm Beach and Orange Counties. Coincidentally, soon to be open mainline barriers at Cypress Creek and Lantana will also serve the majority of all current Turnpike commuter traffi c Since most of the commuter trip links have either origins or destinations in Broward County, three survey locations were selected in this county (S.R. 84/I-595, Sunrise Blvd., and Commercial Blvd.). These entry/exit plazas are also adjacent to each other (3.5-4 miles apart), so survey management could be handled easily. Since Orlando is a major urban area along the Turnpike and also exhibits some affinity for Turnpike commuter traffic, t hree survey locations were selected in Orange/Osceola Counties (Kissimmee/St. Cloud, Orlando South [BeeLine Expressway /U.S. 44 1], and Interstate 4). These entry/exit plazas are also within proximity to each other which again resulted i n easy survey management. Finally, since the highest point of Turnpike commuter traffic is in Dade County, two additional survey locations were selected at the Golden Glades mainline barrier, and along the Homestead Extension of the Flo rida Turnpike (HEFT) at the Tamiami mainline barrier Traffic survey data is not available on the HEFT since it is not on the ticket system, but Tamiami was se lected as the focal point of activity since the HEFT currently carries over 40 percent of the daily Turnpike traffic. Personal Interviews The first part of the Florida Turnpike patron survey involved interviewing patrons at the selected survey locations to determine general characteristics of current Turnpike patrons The sample for this particular survey consisted of 7,154 interviews, all of which were conducted at the eight survey locat ions identified previously. A copy of the interview instrument is included in the Appendix Interviews were conducted from June 26 to June 29, 1990 On each day of the study between the hours of 6:30a.m. and 9:30a.m. and again between 10:30 a.m. and 12:30 a.m., patrons were interviewed in their vehicles while waiting in a queue at a toll plaza. The interview consisted of two observations by the interviewer and three questions asked directly of toll patrons All observations and questions were set up in a closed-ended format. Sampling error for the interview survey is estimated at + I -4 percent. 73


A summary of the observations and questions is provided below: vehicle type vehicle occupancy Are you a Florida resident? What is the purpose of your trip? How many days per week do you travel on this facility? The combined interview results for all survey locations are illustrated graphically in Figure 19. A majority of respondents using the selected toll plazas reside in Florida, as 95 percent indicated they are Florida residents. The passenger car is the principal mode of transportation for those interviewed, as 82 percent of the respondents were traveling in automobiles. In addition, 13 percent were riding in a single-unit truck, followed by o nly 4 percent in semi-trailers. Approximately 77 percent of the respondents were traveling alone while 16 percent had a single passenger. Another 4 percent were traveling with three patrons in the vehicle. A majority of those in tervi ewed (60 percent) indicated that they travel the turnpike five or more days per week. Sixty-two percent of respondents were using the turnpike to commute to work while an additional 18 percent were using the turnpike for some type of company business. 74


FIGURE 19 Personal Interview Results Veh i c l e Type All Survey Locations O ther 1.001 SQmiT ra uor / ., .... U nit T 1 2 .7 $ Are you a Florida Resident? All Survey Locations Y e s 95.M; How many days/week do you travel on this fac ility? All Survey Locations day/week 1 7 31fa d a y twoOk 4 ,001 5 4 d a y s / 'M)(tk 4.6$ 7 5 Vehicle Occupancy All Survey Locations Tht&G 4 ,2$ F ivo or more 1 .2$ What is your trip purpose ? All Survey Locations 1 8 ,0$


Mail-Back Survey The second part of the Florida Turnpike patron survey involved a mail-back questionnaire. A total of 10,400 mail-back surveys were handed out at the same locations and at the same times as where the interviews were conducted A total of 2,088 surveys were returned, resulting in a response rate of just over 20 percent. A response rate of 20 percent or greater is believed to be adequate and therefore increases the comfort level of the survey results significantly. This relatively high response rate can be attributed to a number of reasons including handing the mail-backs directly to potential respondents (as opposed to a direct mailing) as well as respondents having a genuine interest in the A VI concept The mail back survey was comprised of 13 questions. The purpose of th i s survey was to determine patron perception towards the A VI concept and its related operational aspects as well as to develop a profile of the typical turnpike user of those facilities surveyed. Just as in the interview, all questions were developed in a closed-ended format Sampling error for the mai l-back survey is estimated to be + /-4 percent. Mail-back surveys were distributed to turnpike patrons while they waited in a queue at a toll plaza. The survey focused on a number of important issues including willingness to pay for the convenience of A VI, method of payment, trip purpose, trip frequency, and other demographic characteristics A copy of the mail-back survey instrument is included in the appendix. Nearly 83 percent of the respondents were commuting to or from work. This is significantly greater than the pro portion of commuters participating in the personal interviews (approximately 62 percent) The personal interviews provide a better representation of the typical turnpike user s in ce it is more of a random sample of patrons at the survey locations This differs from the mail-back survey in that those who received a mail-back form could choose not to participate in the survey. A disi n terested patron is much more likely to quickly answer five quest i ons in the personal interview than to complete a thirteen question survey. It is expected that, since the frequent user (commuter) has the most to gain from an A VI system, the commuter is more likely to complete a mai l -back survey. 76


Even more significant is the number of job-related trips that were indicated in the mail-back survey including the commute to and from work and other businessrelated trips These proportions are presented by plaza location in Figure 20. FIGURE20 Percent of Job-Related Trips by Plaza Tamiami Golden Gladea SR 84 Commercial Sunrise 1-4 Orlando South Kiaa./St. Cloud 0'!(, 20'!(, 40'!(, 60'!(, 80'!(, 00'!!. Fifty-eight percent of the respondents were male, while 42 percent were female. Most of the respondents (80 percent) ranged from 25 years of age to 54 years of age and a majority o f those responding considered their jobs to be professional (64 percent), clerical (13 percent), and technical (10 percent) positions. Finally, 66 percent of the respondents indicated a household income between $25,000 and $75,000 annually. Tables 5 through 15 summarize the responses to the mail-back survey questions. 77


TABLE 5 Question #I, Mail-Back Survey Would you be willlag to pay for tbe coaveolen

TABLES Question #4, Mail-Back Survey Are you o Florldo resident? Yes 99.0% No LO% TABLE 9 Question #S, Mail-Back Survey What ,.,.. the purpose of tbe trip ou wbieb you received this survey? (check ooly ooe) Work Commute 82.1% Company Business 10.1% Social/Recreation 3.0% School 1.2% Vacation 0.9% Shopping 0.7% Medical/Dental 0.7% Other 0.7% 79


TABLE 10 Question #6, Mail-Back Survey Hoor maay days per wodt do you travel oo this facility? (c:!Jec:k oalr oae) less than 1 day per week 4.9% 1 day per week 2.4% 2 days per week 3.2% 3 days per week 4.7% 4 days per week 4.9% 5 days per week 54.6% greater than 5 days per week 25.3% TABLE 11 Question #7, Mail-Back Survey Gender Male 56.7% Female 42.4% TABLE 12 Question #8, Mail-Back Survey How many vehidts are In your household? (cbtc:k ooly ooe) One 19 .5% Two 57.5% Three or more 23.0% 80


TABLE 13 Question II', MaU-Back Survey Wlaat C8UIO"l' bt$t dctcriboo your occupatl o ll1 (check ODiy 0"") Professiooal 63.5% T ecllllical 10.0% Skilled Labor 4.7% Clerical 13.4% Otbcr 8.2% TABLE 14 Question # 10 Mall-Ba c k Survey What Is your.,.? ( tbeck ollly ooe ) 18 years or le ss 0. 4% 19-24years 6.6% 2S-34years 31. 0 % 35-44 years 30.4% 4.5-54 years 1 9 .1% 55-59 years SS% 3.8% 65 years or m ore 3.1 % 8 1


TABLE 15 Question 11, Mail-Back Survey What is your aooual bousehold iacome? (check oaly ooe) Less than $10,000 2.5% $10,000-$25,000 11.9% $25,001-$50,000 39.6% $50,001-$75,000 28.0% $75,001-$100,000 10.7% G reater than $100,000 7.2% Focus Groups The third part of the Florida Turnpike patron survey was the focus groups. Two focus groups were assembled, including one representing commercial users and one representing the private user (commuters). The purpose of organizing these focus groups was to develop a more-detailed profile of the characteristics and perceptions of the commercial and commuter patrons This survey method permits t he respondent to have a much better understanding of the concept before be/she expresses opinions and perceptions related to A VI in general as well as issues related to the actual implementation of such a system on the Florida Turnpike. Both focus groups were conducted on July 13, 1990. The commercial user focus group met from 9:00 a.m. to 11:30 a.m. and was comprised of eight individuals, representing a variety of commercial interests in the South Florida area. A list of these commercial interests is provided in the Appendix. The initial reaction of the g r oup was extremely positive. All members viewed A VI as potentially improving travel on the Turnp ike significantly. Members of the commercial user focus group voiced strong enthusiasm for tbe potential of A VI to assist i.n the tracking of vehicles. This could be accomplished by providing the commercial user with an itemized statement at the end of each month showing where and when each vehicle passed through specific plazas along the 82


Turnpike. Participants believed strongly that the ability to track vehicles would greatly improve the accounting process for private businesses as well as for the Turnpike An A VI system was deemed particularly superior to the existing credit card In addition, it was believed that A VI could potentially be instrumental in reducing travel time on the Turnpike. A VI is believed to a faster and more efficient method for moving traffic through toll plazas. According to many of the participants, this would also result in reduced fuel costs. Finally, participants indicated that A VI would also save money by reducing the losses associated with giving toll money to drivers and the potential falsification of receipts. In the group preferred a pre-payment system over a post-payment system. Although all participants agreed that a post-payment system is preferred from a business standpoint, most of the group recognized that it would not be advisable for the FDOT to become a creditor. Participants believed that implementing a post-payment system would eventually result in money collection problems and would ultimately result in inflating the price of A VI. Alternatively, there were some arguments supporting a post-payment system. One individual indicated that billing mistakes could be corrected more easily in a post-payment system. The group did recognize that FOOT would enjoy the benefits of interest as a result of the pre-payment system and believed that this could help offset the cost of implementing AVI. Focus group participants indicated a strong preference for dedicated lanes to be made available for A VI-users only. Also, there was strong opposition to any type of barrier in the lane such as gates. Perception indicated that barriers slow down traffic. However, it was mentioned that having no gate would also promote dishonesty. While the majority of the group preferred dedicated lanes, it was clear that a mixed-use lane system would be used if it resulted in better tracking of trucks and in resolving accounting problems. Most of the participants argued in favor of A VI users receiving a discount on the basis that both A VI users and non A VI users benefit from the use of A VI. Consequently, A VI users should receive a discount since they contribute to the benefits that are enjoyed as a result of A VI use. It was also believed that discounts should be offered to encourage A VI use, particularly when A VI is initially offered on the Turnpike. Participants indicated that the discount may be needed if A VI is implemented under a mixed-use lane concept. Although participants indicated a preference for discounts, they did not demonstrate a high degree of 83


resistance to pay for the use of A VI. The perc e ived benefits seemed to offset any concerns about reasonable costs. In general, the focus group preferred an A VI system that employs r adio frequency technology. Th i s was brought about by the belief that radio frequency technology is more reliable than optical (bar code) technology. However, there was still some interest in the bar code technology if it resulted in significantly lower costs without significant loss in reliability Also participants indicated a preference to lease a transponder rather than purchase one. Users did not want the responsibility of having to maintain the transponder if it were to malfunction. In addition, it would be less costly to discontinue participation in the system if dissatisfied. The second focus group was comprised of 12 individuals representing the private user (commuter). All members of this group lived in the Miami/Ft Lauderdale area and commute on the Turnpike at least five days per week. This focus group was conducted from 1:00 p.m. to 3:30 p.m The results of the commuter focus group were very similar to that of the commercial focus group, with some minor variations. Most reacted positively to the concept of A VI; however, this group seemed to be much more sensitive to the cost of participating in the A VI system. Nearly all of the participants expressed a strong desire to have a pre-payment system. It was believed such a payment system would be more convenient for the user as well as for FDOT. This group also believed that it would be a mis t ake for FOOT to become involved in the collection business. There was some concern about a pre-payment system that included automatic withdrawals from a user's credit card Some of the participants felt that they might lose some control over their credit card account. Regardless of the payment method, participants expressed an interest in receiving itemized statements at the end of each month. It was believed that i t emization would lead to a higher level of user comfort. I t was even suggested that an itemized bill would not need to be a requirement but should be available for an extra charge Focus group members expressed st r ong s e nt i ment about receiving discounts if they participated in the A VI program, especially with a pre-payment program where FOOT 84


would enjoy the benefits of interest earned on the pre-paid money In addition, since everyone would benefit from an A VI system, A VI users should be rewarded for their participation in the program. Discounts would also provide a good method for encouraging participation. A majority of participants were reluctant to pay any type of premium to use A VI. For example, most believed that having to pay a one-time fee of S50 to join the program with a $2 per month service charge was too costly. One or two participants believed that it was FOOT's responsibility to make improvements to the Turnpike. Therefore, if A VI would help FOOT accomplish this goal, then FOOT should provide this service at no additional charge All participants indicated a preference for dedicated A VI lanes in order to receive maximum benefit from an A VI system. However, most participants perceived that benefits are still realized in a mixed-use lane system. Just as in the commercial group, private users indicated a preference for radio frequency technology. However, there was some interest in the optical (bar code) technology, depending upon the cost of the radio frequency technology. Participants also preferred to lease the transponder, particularly if the radio frequency technology were used In general, participants believed it to be more convenient to lease rather than own a vehicle transponder. Finally, focus group participants expressed a strong desire for a universal A VI system to be implemented on toll facilities statewide. A VI users could then travel throughout the state and still be able to take advantage of the benefits that an A VI system can offer. Summaey and Conclusions In summary, five areas of survey research were considered as part of Phase II of the project including surveys of existing A VI users, a summary of other A VI-related surveys, Florida Turnpike patron interviews, Florida Turnpike patron mail-back survey, and Florida Turnpike patron focus groups. The results of this survey research provide information which assists in assessing the market potential of A VI on Florida's Turnpike. 85


MARKET POTENTIAL Turnpike Characteristics Existing daily traffic on the total Florida Turnpike system is on the order of 450,000, according to the 1988 Traffic Survey Report compiled by URS/Coverdale & Colpitts. Also, according to this report, about 53 percent of this daily traffic is on the Turnpike Mainline, 43 percent is on the HEFT, and 4 percent on the Bee Line Expressway West. Approximately 49 percent of the 1988 total daily Turnpike system traffic consisted of commuter type (five or more day/week user)traffic. Accord in g to the 1989 Turnpike Bond Indenture, during the previous 10 years, traffic on the Turnpike system (mainline, HEFI, and Bee Line) has increased about 132 percent, and for the next 10 years, Turnpi ke system traffic is expected to increase another 82 percent. Therefore, assuming travel characteristics on the Turnpike system do not change significantly over the next 10 years, close to 400,000 daily commuters could be expected to utilize the Turnpike system on a daily basis by the year 2000. Traffic growth currently forecasted for the Turnpike for the year 2015 will at least double and in some locations almost triple 1989 volumes. The greatest growth in traffic is expected to occur in Palm Beach County, between Palm Beac h Gardens and Delray Beach (38,000 to 98,000 vehicles/day). Given the trends highlighted above, it appears practical that the first priority for implementation of A VI should be considered along the HEFT and mainline Turnpike, primarily between Palm Beach and Dade counties. The Turnpike characteristics suggest that A VI has market potential if the users respond positively to suc h a service. An assessment of the market potential based on survey results is provided in the next section. Assessment o( Market Potential The Florida Turnpike patron surveys provide data that can result in a general assessment of the market potential o f A VI on the specific sites included in the survey. The surveys 86


establish important factors regarding the profile of current Turnpike users, the estimated level of participation in an A VI system, and perceptions toward various A VI operational characteristics. User Profile As expected, the surveys indicated two major market segments that would need to be penetrated in order to attain sufficient participation in an A VI They include the everyday commuter and the frequent commercial user. Approximately 80 percent of all interview respondents (n=7,154) indicated the purpose of their trip to be oriented to a work commute or related to company business i n general. This finding is supported in the mail-back survey, where nearly 93 percent of respondents (n=2,088) indicated their trip was related to these same two types of trips. Additionally, it is expected that these trip types are more likely to be taken by Florida residents, a fact supported by both surveys since 95 percent of interview respondents and 99 percent of mail-back respondents indicated this was the case. This is particularly significant when you consider that the 7,154 interviews were random in the sense that there was an equal opportunity for Florida residents and non-Florida residents to be questioned. However, one would expect to receive a greater proportion of mail-backs from Florida residents since they travel the facility more frequently and therefore have a genuine self interest in a service that may improve their trip on the Turnpike. Most important is th e response to both surveys concerning the number of days per week that they travel on the facility at these specific Turnpike locations. Sixty-one percent of interview respondents and 80 percent of mail-back respondents travel these locations five or more days per week. The frequent Turnpike user is obviously the most important segment to penetrate since they have the most to gain from such a service. The patron interviews indicated that the typical Turnpike user normally travels the se lecte d Turnpike locations in a passenger car (82 percent) or a single unit truck (13 percent). The third most frequent vehicle type is the semi-trailer at nearly four percent. Vehicle type was not considered in the mailback survey. 87


Level of Participation It is expected that the level of participation will vary depending on the operational characteristics of the system. However, it does appear that cost is the most important issue concerning participation in a A VI system. Specifically, respondents were asked i n the mail-back survey whether they would be willing to pay for the convenience of A VI service under different cost scenarios such as discounted toll rates, normal toll rates, or some type of premium toll rate. As expected, perceived level of participation was highest under the discounted toll rate scenario and lowest under the premium toll rate scenario. Over 90 percent of respondents indicated a desire to use A VI under the discount scenario. If toll rates remained the same, approximately 67 percent indicate they would still participate in the A VI system. Finally, the suggestion of a premium charge resulted in a dramatic decline in perceived participation, as only 16 percent indicated they would use A VI under the highest cost scenario. It is important to recognize that this perceived willingness to participate is limited by two major considerations. First, the question mentioned the magnitude of the toll rates only and did not consider the cost of the transponder or the other operational characteristics of the system. Second, analysts should always consider hypothetical survey results with caution as it has been discovered that hypothetical responses are not always borne out in reality. These two considerations play an important role in determining the willingness of patrons to participate in an A VI system and were considered more closely in the focus groups. Perceptions of AVI Operational Characteristic$ -The mail-back survey was also designed to establish attitudes toward some basic operating characteristics of an A VI system. Specifically, respondents were questioned regarding their preferred method of payment for A VI. A majority (60 percent) indicated a preference to pre-pay for A VI with cash or check whil e a small percentage (7 percent) indicated a preference to pre-pay by credit A third alternative was a charge against a credit card which could be perceived as either a direct charge for each use on the Turnpike or a post payment with the credit card after accruing a specified toll amount. Over 33 percent of respondents indicated a preference for this third payment option. These perceptions are particularly interesting when one considers that over 90 percent of mail-back respondents indicated having a major credit card; however, the majority still preferred a pre-payment system with cash or check. 88


Various A VI operational characteristics were explored more thoroughly in the focus group surveys. The preferred characteristics varied in some aspects from the commuter focus group to the commercial user focus group; ho wever both indicated specific benefits of various configurations of A VI systems. Both groups reacted positively after the concept of A VI was introduced and described to them. likewise, the commuters and commercial representatives supported the following: pre-payment system, discounted fares, i temized statements, dedicated lanes, radio frequency technology, and leased transponders. Most importantly, the commercial user representatives believed that A VI would provide the ability to track vehicles along the turnpike; however they also agreed that the system would reduce travel time. The commuter representatives emphasized time savings and convenience as t he primary benefits of an A VI system; however, the commuters were much more sensitive to any costs involved with a system. Summaa of Market Potential In summary, the user profile and perceived level of participation indicate great market potential for an A VI system at those locations included in the interview and mail-back surveys. These surveys were also supported by the observations made during the commercial and commuter focus groups. However, the focus group results do indicate that many A VI characteristics are preferred over others. Therefore, the specific operational characteristics need to be analyzed closely in order to maximize the actual level of participation. 89


VI. COSTS AND BENEFITS COSTS This section provides a consideration of the capital, operating, and maintenance costs associated with the construction and operation of various Jane types on a toll facility. In particular, cost considerations include Jane construction cost and right-of-way, equipment cost by Jane type, operating and maintenance cost, computer hardware and software cost, and transponder cost for the different technologies. It is important to note that, when the listed equipment is purchased in large quantities, the unit cost will likely decline. Lane Constrvction Cost and Ri&h!=of-Wsy Industry standards suggest that engineering/ construction cost for the average toll lane is approximately $500,000 plus the cost of the booth and equipment placed in the lane. It is important to recognize that this is only an estimate as this cost will vary by region and location throughout the United States. This estimate includes planning, concept design, lane construction, right-of-way, and other associated lane construction costs. Cost of Egujpmj:nl by Lane type Tables 16 to 20 represent line-item costs for various Jane types including manual lane, automatic coin machine, A VI dedicated lane (retrofit), mixed-use lane (A VI and manual), mixed-use lane (A VI and automatic coin). For each line-item, a range of cost is provided that is believed to include the majority of the vendors who provide that particular item in the industry. A mid-point is also calculated for the purpose of establishing a benchmark for the average cost of a particular piece of equipment It is important to recognize that the cost estimates provided are extremely conservative in nature so as to virtually eliminate any possibility of underestimating equipment costs. The equipment depicted under each configuration assumes the maximum use of various equipment components. Some of the equipment would certainly be optional depending upon the specific location and configuration. 90


The cost of the various lane types ranges from approximately $35,000 for an A VI dedicated lane with no other form of payment option to nearly $110,000 for a lane equipped with A VI and an automatic coin machine. For each manual lane, a booth must be installed which is conservatively estimated to cost $30,000. The booth is not included in the line-item costs in the tables below. These mid-point lane cost estimates will be used in the site-specific cost-effectiveness analysis conducted for Tamiami Plaza later in this report. ComPuter Egpipment and Software Cost Included in the consideration of hardware are the plaza computer equipment and the host computer equipmenL In addition, the cost of both the plaza and host software must be identified as well. Figures 21 and 22 illustrate a sample plaza computer configuration and a sample host computer configuration and are each followed by Tables 21 and 22, respectively, which represent the cost for the various components of the configurations. Again, the cost estimates provided are extremely conservative in nature and the configurations include certain equipment components that could perhaps be eliminated depending on the specific location. In particular, dual processors were included for built in redundancy to ensure 24 hour, seven day a week availability resulting in uptime in excess of 99 percent. Dual recording of transaction data on shadow recorded disks is also a feature included as is automatic switchover to the backup computer when necessary. Operating and Maintenance Cost The cost-effectiveness analysis included in this report assumes an $81,000 per lane operating and maintenance cost over the life of the analysis. This 1991 cost represents the average per lane cost based on the OOCEA budgeted expense of $9,310,392 spread over 115 toll lanes (includes eight mainline plazas). Since this per lane cost is based on an untypical ratio of automatic lanes to manual lanes (1:3), it is deemed a conservative per lane operating and maintenance cost for Tamiami Plaza. The rationale for this assumption is developed more thoroughly in the cost-effectiveness analysis of A VI presented in Section VII. 91


-{; '$ '!""1' -.. j 92 ] 8' .8 a r "3 "' <=' a i!, a ::J 2 s 0 -. :a ,g : 1-< s -.!.


TABL AUIOalllic LaDe Equ.ipmeat Coot I ''' 'C<>q .. . "<"'#11:':"......-. ..;_,.,,.:,;;<,.,_, f;f9 ..... !;c;.. . :;..; I t;.. .. 11> i- :., *'**'.-... *'" .. 8 000 ..... $45,000 '" ,. ...... ,. .... l .. :t-t:!1:* . ... : 4b'-:: : "' I> ,.,.,.E,..t-' """' "., ... .. ---><- ............. ... 1..oof CootroU.r --'<'-'*'*ti ;$q\ti .. "'-........ .. : . ; $250 I Exit Gale .. .... LaDe Gole (luo clooUog) -...... Tpflic I t I Caaopy Sipll Ligbl ., ___ ,,,.'lot_ ..... _.__.,._ -Mi',,, """"' : $500 ... "2 400 1 $500 $300 ........... $2,400 --... $300 $1 ,200 ':"'"-*'AF ,.. _ ., r "''" our ....,wa .-;.il>.-".1-.............. .-.,. __ .. dividuol type of cooligunlloolperipberla ___ ,.. .... . "'2 450 :.:. ..t:! tJ;. "' f ... $625 . 1"'..-. $4 250 -..... .._ $200 ____ .__,. e: c 1,450 $1 ,450 -=:t:.-. $1,150 ...... $750


Il "5 a ., .. "' i gj ., ., "' 8 "' ... .. l .8 1 i i i .e-z. "S 1 i u:: .. il a i > < ,11 ...... 0 ->. .. J 1 <"< o; '"' 5 > < z 94

PAGE 100

< ,, \0:,, ., 95 -f h .2

PAGE 101

< ; < .... 'fi .. "'" t: '$, "''* ;. < 96

PAGE 102

:s FIGURE 21 DUAL PLAZA COMPUTER CONFIGURATION 4mm Tape -----..: System CMik ............... .. ': !;liii: :;:!!!H'''' :!i.Y"' .. .. : h :t: :.,._. Data Obk i LANE I e ee jj. lAN E n lil;; ::w; u:: ::iw' IJ::,,, !li ::1; 1 :1 ;: Remot4t Lanes .. --. .,.-r-. . ... m PRIMARY PLAZA I ml sr ..... COMPUTER NETWORK 802.3 Ethernet TEMitUUl Prlnter ....... ... . .. ,! ,t W;W,. ; ;i;i; ,;,;; .. Syslom lllal< .. .< plAZA ... ,lt:lj';:: l::;: ; : !';!:Vi" COMPUTER ''" Prlnlat !!I LANE I ( !!! mmmmlllHiil!}' ,,, Oaaa Olsk Ft.or O!Mic 62. 5 ST Cable 10 Each Lane Conatolw .... ,.:., . . ,.!i" L ocal Lanes

PAGE 103

f 98

PAGE 104

== w 1-f) en { : f j > I I en l 0 ... di """' N N a: r.l w i =: 1-;;;> {,:) ::::> ; l j c. s '"" 'I == '.../} 0 (,) 1-en : 0 ::1: A j!, >i l h I Q f!/ .. \ .. /"\ ., ' U : (;; 'v!]/ -u ' 99 .. I t, J ' l J 3

PAGE 105

-8 TABLE 22 Hoot Compub>r Equipment Cost Eslimalos . :.hi ','"" ""2 .. >.,.. ,., -2 . '->-"11' '< ''<' X.: ,._ "' ">? ' N""""'k .. ;-,:$ .. 2 Terminal Server < ,)I ct>; < ($; 4 12 ....... 2 ;,: . and Prio rer) .. NOTE: L Som<> of lilo idenlilied equipment may be optional for a specific coo6guralioo. 2. Additiooal hoot oompub>r and disk drives included for redundancy.

PAGE 106

Tran sponder Cost The unit price for a radio frequency transponder ranges from $25 to $42 depending upon the technology, the vendor, and the quantity of transponders pu r chased. Alternatively, the u nit price for a bar code deC!11 ranges from each for large quantities (100.000) to each for smaller quantities (2,000 to 3,000) N ois e P o ll utio n Pollution in any form is a great concern to many environmental groups and organizations For example, many state and local governments have vehicle noise standards. In Florida, the State Motor Vehicle Noise Standard for a car or light truck (weighing less than 10,000 pounds and at a distance of 50 feet) cannot exceed 80 decibels The Vehicle Noise graph, Figure 23, shows the relationship between vehicle speed (miles/hour) and noise l evels ( d ecibels). To obtain this grap h the data were procured from the National Reference Energy Mean. The pos i tive slope of the line indicates tbat the noise level is directly proportional to the increasing speed of the vehicle; that is, as the vehicle accelerates at a constant rate the no i se l e vel increases Higher toll p r ocessing speeds associated with A VI might result in overall higher noise levels. Following any A VI implementation, noise levels should be monitored to determine i f mitigation is r equired FIGURE23 Vehicle Noi se By Speed r -----------.0 ------------------------------SOUACE: CMr9Y ........ fP'ecwel ... 01 S.OIIOII 23 CIIR. Part 7T1) 101

PAGE 107

BENEFITS There are numerous benefits that could be realized as a result of the implementation of an A VI system. A list of these benefits along with a brief description is provided in this section Throughput Emciency The imp l ementation of A VI will increase toll lane capacity, thereby reducing toll processing time and queue lengths at toll plazas Time savings, or opportunity costs, are reflected in the cost-effectiveness analysis portion of this report as a component of the vehicle delay cost differences. The FOOT D-QUE UE Plaza Simulator Model was utilized to determine ideal plaza lane configurations, as previously mentioned in this report. Output from this model quantifies: maximum queues (in vehicles), average queues (in veh i cles), and average delay (in seconds per vehicle) for each 15-minute i n crement during the peak hour. Input requirements include approach volumes in 15-minute increments, number of approach lanes, number of plaza lanes, and plaza lane configuration by lane type (capacity). The maximum number of plaza lanes in the peak direction that can be simulated is seven; therefore, plazas with a greater number of lanes have approach volumes prorated (as necessary) The appendix of this report contains sample outputs of D-QUEUE as described above. Total delay in the peak hour can be estimated bY multiplying the number of vehicles bY the average delay for each 15-minute period. Maximum queues and average queues in the peak hour can also be estimated from this output Queue lengths can be determined by multiplying the queue (given in vehicles) by th e average l e ngth of a vehicle (20-25 feet) Fue l Sayings Data gathered from NCHRP Report #111 "Running Cost of Vehicles as Affected by Highway Design and Traffic," indicates a basic relationship between vehicle speeds and fuel 102

PAGE 108

consumption Figure 24 illus trates the relationship between vehicle speed (miles/hour) and the amount of vehicle fuel consumpt ion (gallons/mile). This relationship depicts a pattern t hat fuel consumptio n decreases as the vehicle begins to accelerate, then increases after reaching speeds over 30 miles/hour. With respect to mathematical formulation, the pattern indicates that the speed is a function of gallons/mile exponentially Fuel consumption with respect to A VI implementation will have a positive effect on fuel savings. As gas prices rise, fuel consumption will become a major priority among patrons of the Turnpike. Whenever a vehicle goes through a conventiona l toll plaza, the vehicle must slow down to pay t he toll, which causes the vehic l es to consume more gasoline at slower speeds As A VI is implemented, vehicles will proceed through the plaza at much more efficient speeds, thus reducing overall fuel consumption. It should also be noted that for express A VI lanes (55mph}, fuel consumption would again compare with the conventional plaza lane s. The ideal speed for fuel consumption is approximately 30mph. The ideal speed and lane configuration of A VI for fuel consumption would be a dedicated A VI lane (15mph average, 30mph abso lut e maximum). Fuel savings can be quantified in terms of reduced idling costs, and are included in the cost-effectiveness analysis portion of t hi s report. FIGURE 24 Fuel Consumption Of Passenger Vehicles G 0 .05 r i I 0.04 r------. I ---. .. I 0.02 f------.... ' I 0.01 r----------.. -- -. --------. --' i SOUR CE 0 m m 30 40 so so ro Speed (mph) NCHRP Report #111 RunninG Coat of Vehlc fee u Affected by H ighway OeiQn TraffiC 103

PAGE 109

Air Pollution According to the Environmental Protection Agency (EPA), motor vehicles account for only a small share of most of the nation's air pollutants. However, motor vehicles do account for almost 60 percent of the nation's carbon monoxide emissions Tests run on different types of vehicle exhaust emissions are usually run simultaneously with fuel consumption tests. The three emissions tested for by the EPA MOBILE4 Emission Factor Model were carbon monoxide (CO), non-methane hydrocarbons (NMHC), and nitrous oxides (NO,.). These relationships are illustrated in Figure 25. By comparison, all three graphs indicate that the emissions decrease as the vehicle accelerates and continues traveling at a constant speed Compared to a conventional toll facility, an A VI facility may have a positive affect on air quality. Since the flow of traffic is enhanced and speed is increased by implementing the A VI project, it would result in less pollution per vehicle. 104

PAGE 110

FIGUR.E l5 Vehicle Emlsslo FldOn Carbon Monoxide tor ... t20 --. -tOO ----- --- --. eo --------........ ---::::::.. :;:::: 4 -. ----.. --0 I , 0 M '0 10 1Spood (mph) --oo Non-Methane Hydrocarbons --------------- r-. ------- 0 0 .. _..,_ ... ________ -Oxides of Nitrogen u __ :::::::_ .. -___ --------1 :L_--_ __ .. .. _ _ __ ____ __ .. -__. 1 -------0 .. 20 ... o .. .. Speod ( mph) 105

PAGE 111

Security A23inst Loss Of Funds Most revenue funds lost by a toll collection system are due to human error or fraud. The automation of systems resulting from the implementation of A VI increases the security against error or fraud. Processes once handled manually, such as reconciliations, axle classifications, transaction report generations, and toll evasion enforcement, can be accomplished through the system's computer and video cameras. The ability to generate a clear audit trail of revenue funds through an automated and networked A VI system is a major security benefit for those revenues collected from A VI-equipped vehicles. Improved Accountability A computerized A VI accounting system leads to the benefit of improved accountability for A VI transactions. First, the system eliminates errors caused by the manual processing of accounting and operational reports. SecondJy, the system provides improved statistical . traffic reports, since all A VI-equipped vehicles passing through the toll system can be accounted for electronically. Also, by computerizing many functions of a toll system, audit control points are easier to implement and maintain. Payment Alternative ror Patrons A VI allows toll patrons more flexibility and convenience in paying tolls. The opening of pre-payment accounts eliminates the need for patrons to be concerned with having cash ready for each toll plaza passage. Prepayment also provides patrons the flexibility of paying tolls by cash, check, credit card, or electronic funds transfer. In addition, A VI reduces the need for the handling of hard currency by toll system operators Furthermore, commercial users of A VI are given a faster and more reliable tracking system of when and where their vehicles use the toll system. 106

PAGE 112

Reduced Construction Costs There are three alternatives that an agency can select from when an existing plaza is at or near capacity: (1) expand the number of Janes at the p laza (2) increase traffic throughput through the existing configuration (3) accept a lower level of service Accepting a lower level of service is unlikely due to the resulting loss of revenue and safety that would likely occur. Expansion of the plaza has typically been the solution implemented most often in toll facilities across the United States. However, expansion involves significant capital outlays (approximately $500,000, plus booth and equipment costs) per lane. In addition, in many cases expansion may be impossible where there is no right-of-way available A third alternative is the implementation of A VI on the existing configuration, which can significantly increase the capacity of the existing lanes. This inc rease in throughput can result in a significant reduction in necessary lane construction that may otherwise be needed. Perhaps an even more feasible alternative is to implement A VI in conjunction with some degree of expansion. Reduced Ooernlln& Cost The operating costs associated with a typical manual lane include the costs of maintaining the lane equipment, collector and supervisor salaries, auditing functions, banking and cash handling, and lighting and environmentally controlling the booth that houses the toll collector. Similarly, the costs of operating an automatic coin machine lane includes those associated with maintaining the lane equipment, supervisory salaries, auditing, and banking and cash handling. For each manual lane eliminated as a result of A VI. these particular operating costs will be significantly reduced. The collection costs for an A VI system include those associated with maintaining the A VI equipment, issuing tags, and servicing accounts In addition, the system operates 24 hours a day. These collection costs will also be partially offset by the monies that are collected 107

PAGE 113

as part of the sale and/or lease of the tags and the interest realized on pre-payments and deposits. Potential for Increased Revenue By offering an additional payment alternative, the Turnpike may encourage users of an existing parallel facility to convert to using the Turnpike. This is particularly true in a location where an existing parallel facility is undergoing construction. The increase in Turnpike patronage could result in an increase in total revenue for the Turnpike. Increase in Patron Satisfaction and Lqyaltl By offering the patrons an additional payment alternative that makes their travel quicker and more convenient, the Turnpike can expect the typical Turnpike user to be much more satisfied with the facility as a whole. In addition, once a Turnpike user begins participating in the A VI system, that user will be much more likely to use the Turnpike as opposed to an alternate parallel facility since they now have a tag attached to their vehicle. Other Benents Several other non-quantifiable benefits can also be attributed to A VI implementation. These would include non-user benefits, non-peak hour A VI user benefits, potential for variable (or congestion) pricing of tolls, and shared A VI commuting costs with ridesharing. Non-A VI users would benefit to some degree by less overall plaza congestion, particularly in mixeduse lanes. Since differences in vehicle delay costs bave been considered for the peak direction, peak hour analysis, additional benefits can also be expected in the other direction and in other time periods of the day. As other adjacent roadway facilities become congested, variable toll pricing (reducing normal toll rates) can be used as an incentive for patronage to divert their travel to the uncongested or less congested toll facility. A VI can also be an incentive for ridesharing since A VI costs could be shared. 108

PAGE 114

COST ESTIMATES FOR AVI IMPLEMENTATION Given the level of detail of this report, it is not possible to determine site-specific A VI implementation costs. However, order-of-ma&Jlitude AVI costs can be estimated utilizing cost data contained in the previous section of this report and the existing number of toll lanes by segment (assuming fllll A VI implementation for purposes of discussion only). For simplicity, A VI costs in this estimate include only A VI lane equipment, 100 percent of A VI computer hardware and software (plaza and host) costs, and five percent contingency. The first priority segment, that portion of the Turnpike system south of lantana including the HEFT, presently contains approximately 121 toll lanes and five mainline plazas. Total estimated order-of-magnitude A VI costs for this segment would be about $6.5 million. These costs assume full A VI implementation, which may not be warranted (but assumed for this estimate only). The second priority segment, that portion of the Turnpike system between and i ncluding the Orlando West interchange and the Kissimmee/St. Cloud interchange, presently contains approximately 15 toll lanes and no mainline plazas. The estimated orderof-magnitude A VI costs for this segment would be about $1.4 million. These costs assume full A VI implementation, which may not be warranted (but assumed for th i s estimate only). 109

PAGE 115

VII. COST-EFFECTIVENESS ANALYSIS OF AVI INTRODUcnON This section was prepared to document the methodology and results of a cost effectiveness analysis of implementing A VI at the plaza level. Tamiami Plaza, which is located on the HEFT, was selected for this analysis. The analysis compares the costs of implementing various improvements under four scenarios with the road user benefits that would be realized under these same four scenarios. Costs include lane construction, lane equipment. computer hardware and software, operating and maintenance, transponder/decal costs, and contingency costs. Benefits include the reduction in road user costs that are expected to result from the improvements made under the four different scenarios. The cost and benefit components are described in detail later in this section. It should be recognized that throughout the analysis, al l numbers used are deemed conservative according to industry standards. The intent of this conservative strategy was to virtually elintinate the possibility of underestimating costs and overestimating benefits. Four scenarios were considered, including the currently proposed construction alternative and three A VI configurations that represeni three levels of A VI participation (10 percent, 30 percent, 50 percent). Each scenario is compared with the baseline scenario, which is the "no -build do-nothing" alternative. The analysis begins in 1994 and is conducted through the year 2015 when the proposed construction for Tamiami Plaza is expected to be complete. The Turnpike has set an acceptable queue length criterion as being 300 feet. Therefore, the proposed construction through the year 2015 is designed to preserve this criterion. Ukewise, the three A VI configurations are also designed to preserve this criterion. The scenarios were developed through an ite rative process using D-QUEUE, a plaza simulation model that has been discussed previously in this report. The ideal number and configuration of lanes was developed under each A VI scenario to meet the specified criterion. The net present value, benefit-cost ratio, and internal rate o f return is computed for each alternative and will determine the economic desirability of each scenario as compared to the no-build, do-nothing alternative. In addition, the indicators fo r each scenario will be compared to determine the feasibility and rank of the investment strategies. 110

PAGE 116

METIIODOLOGY The guidelines for conducting the cost effectiveness analysis were taken from the American Association of State Highway and Transportation Officials' (AASHTO) publication, A Manual on User Benefit Analys i s of and Bus-Transit Improvements. 1977. The AASHTO manual provides a nationally-accepted methodology for transponation project cost effectiveness analysis. The manual emphasizes road user benefits and agency costs Secondary costs and benefits, which are difficult or impossible to quantify, are not included. Three major indicators of economic feasibility are calculated, each of which is defined below. Net Present Value (NPV) the difference between the present value of total p e riodic benefits and the present value of total periodic costs If the NPV associated with an alternative is greater than zero the alternative is economically justified Benefit/Cost ( B/C) Ratio the ratio of the net present value of total periodic benefits to the net present value of the total periodic costs. The project alternative can be economically justified if the B/C ratio is greater than 1.0 Internal Rate of Return (IRR) the discount rate which equates the present value of a project s expected cash inflows to the present value of the projects expected costs That is, the IRR is equal to the discount rate for which NPV = 0 and B/C ratio 1.0. The project alternative can be economically justified if the IRR is greater than the discount rate used in the analysis. D ETERM I NATION OF USER BENEFI T S Consistent with the AASHTO total road user costs were calculated for the "no build, do nothing alternative and for the four improvement scenarios User benefits are .calculated by subtract i ng total road user costs for each scenario from the total road user costs of the "no-build, do-nothing" alternative If the d i fference represents a reduction in 111

PAGE 117

road user costs, then that value constitutes a benefit and can be used as such in the calculation of net present value (NPV) and benefit/cost (B/C) ratio. There are three major components of total road user cost: vehicle stopping costs, vehicle delay costs, and vehicle idling costs. Vehicle stopping costs were not included as it was determined that this cost would be the same for the no-build alternative as for the four improvement alternatives. The vehicle delay costs and the vehicle idling costs were calculated as prescribed in the AASHTO manual for establishing intersection delay costs. Vehicle delay costs and vehicle idling costs were calculated and escalated to 1990 dollars using the Consumer Price Index (CPI). The combined cost of delay and idling is $0.00249 per second of delay per vehicle. Therefore, total road user costs for each year are calculated as in the following equation: Total Road User Costs = ADVD x #VEH x COST where: ADVD = average daily vehicle delay in the peak hour #VEH = number of vehicles in the peak hour COST = road user cost per second of vehicle delay The resulting road user costs for the no-build alternative and the improvement alternatives represent the costs for the peak hour, peak directio n only. These costs are provided in Table 23. User benefits are calculated by subtracting the road user costs of the improvement alternatives from the road user costs of the no-build alternative. The user benefits are provided in Table 24. However, because the calculated road user costs were representative of the peak hour, peak direction only, the resulting road user benefits are also representative of the peak hour, peak direction only. A more realistic representation of user benefits would include the road user benefits realized throughout the day. However, data limitations prevented the calculation of road user costs and benefits for the entire day. For this reason, the road user benefits calculated for the peak hour, peak direction were doubled to account for the peak hour in the opposite direction. The resulting benefits are still extremely conservative since they do not include the benefits that would be realized in the 112

PAGE 118

shoulders of the peak and in the off peak hours. The adjusted user benefits for each improvement alternative are provided in Table 25. The user benefit table indicates that benefits are virtually the same under each of the alternatives. This is explained by the original methodology which requires each alternative to maintain a standard criterion for a queue length of 300 feet. Therefore, each A VI alternative is configured in a manner that will maintain the same criterion as in the construction alternative. As a result, all four improvement alternatives would be equally beneficial to patrons as they pass through the plaza. The true benefits ultimately result from the reduction in lane construction and lane equipment costs that could potentially result from the implementation of A VI 113

PAGE 119

TABLE23 Total Ro a d Use r Costs for NoBuild and Impro vement Alt erna tiv es Year No-B uild C oatructioa 10%AVI 30%AVI 50%AVI 1994 $1,228,465 $22,978 $ 20,359 $18,289 $16,2 1 9 1995 $ 1 283,716 $24,109 $21,491 $19,421 $17,351 1996 $1, 338,966 $25,240 $22,622 $20,552 $18,482 1997 $1,394,217 $26,371 $23,754 $21,684 $19, 614 1 998 $1,449 ,467 $27,502 $24,885 $22,815 $20,745 1 999 $1, 504,717 $28,634 $26 ,017 $23,947 $21,877 2 000 $1,559 ,968 $29 ,765 $27,148 $25,078 $23,008 2001 $ 1 ,615 ,218 $30 896 $28,280 $26,2 1 0 $24,140 2002 $1,670,469 $32,027 $29,411 $27,341 $25,271 2003 $1,725 ,719 $33,158 $30,543 $28,473 $26,403 2004 $1,780,970 $34,293 $3 1 ,674 $29,604 $27,534 2 005 $1,836,220 $29,237 $25,905 $23,271 $20,637 2006 $1,891,470 $30,790 $27, 458 $24,824 $22,190 2007 $1,946,721 $32,343 $29,012 $26,378 $23,744 2 008 S2,001, m $33,896 $30,565 $ 27,931 $25,297 2009 $2,057,222 $35,449 $32,119 $29,48 5 $26,851 2010 $2, 1 12,472 $37,002 $33,672 $31,038 $28,404 2011 $2,167,723 $38,555 $35,226 $32,592 $29,958 2012 $2,222,973 $40 108 $36,779 $34 ,145 $31,511 2013 $2,278,224 541,661 $38,333 $35,699 $33, 065 2014 $2,333 47 4 543,2 1 8 $39,886 $37,252 $34,618 2015 $2,33 3 ,474 $34,393 $30,473 $27,375 $24,277 T otal $39 ,733,836 5 711,625 $645,61 0 $593,402 $541,194 11 4

PAGE 120

TABLE24 Ro a d User Benents, Peak Hour, Peak Direc tion Y ear CoDStrucdoa lOCk AVI 30%AVI SO'li>AVI 1994 $1,205,487 $1,208 ,106 $1,210,176 $1,212,2A6 1 995 $1,259,607 $1,262,225 $1,264,295 $1,266,365 1 996 $1,313,726 $1,316,344 $1,318,414 $1,320 484 1997 $1,367,846 $1,370,463 $1,372,533 $1,374,603 1!198 $1,421,965 $1,424,582 $1,426,652 $1,428,722 1 999 $1,476,083 $1,478,701 $1,480,771 $1,482,841 2000 $1,530,203 $1,532,820 $1,534,890 $1,536,960 2001 $1,584,322 $1,586,939 $1,589,009 $1,591,079 2002 $1,638,442 $1,641,058 $1,643,128 $1,645,198 2003 $1,692,561 $1,695,177 $1,697,247 $1,699,317 2004 $1,746,677 $1,749,296 $1,751,366 $1,753,436 zoos $1,806,983 $1,810,315 $1,812,949 $1,815,583 l006 $1,860,680 $1,864,012 $1,866,646 $1,869,280 2007 $1,914,378 $1,917,709 $1,920,343 $1,922,977 2008 $1,968,075 $1,971,406 $1,974,040 $1,976,674 2009 $2,021,773 $2,025,103 $2,027,737 $2,030,371 2 0 1 0 $2,075,470 $2,078,800 $2,081,434 $2,084,068 2011 $2,129,168 $2,132,497 $2,135,131 $2,137,765 2012 $2,182,865 $2,186,194 $2,188,828 $2,191, 462 2013 $2,236,563 $2,239,891 $2,242,525 $2,245,159 20 1 4 $2,290 256 $2,293.588 $2,296,222 $2,298,856 2015 $2,299,081 $2,303,001 $2,3()6,099 $2,309,197 Total $39,022,211 $39,088,226 $39,140,434 $39,192,642 115

PAGE 121

TABLE25 Road User Benefi t s, Bo t h P ea k H ours y.,... Cou$trud:iOG 10%AVI 30%AVI 50%AVI 1994 $2,410,974 $2,4 1 6,212 $2,420,352 $2,424,492 1995 $2,519,213 $2,524,450 $2,528,590 $2,532,730 1 996 $2,627,452 $2,632,688 $2,636,828 $2,640 968 1997 $2,735,69 1 $2,740,926 $2,749,206 1 993 $2,843,930 $2,849,164 $2,853,304 $2,857,444 1999 $2,952,167 $2,957,402 $2,961,542 $2,965,682 zooo $3,060,406 $3,065,640 $3 069,780 $3,073,920 Z001 $3,168,645 $3,173,878 $3 178,018 $3,182,158 2001 $3,276,883 $3, 282,115 $3,286,255 $3,290,395 2003 $3,385,122 $3,390,353 $3,394,493 $3,398,633 2 004 $3,493,353 $3,498,591 $3,502,731 $3,506,871 Z005 $3,613,966 $3, 620,630 $3,625,898 $3,631,166 l006 $3,721,361 $3,728,024 $3,733 292 $3,738,560 2007 $3,828,756 $3, 835,418 $3,840,686 $3,845,954 2008 $3,936,151 $3,942,812 $3,948,080 $3,953,348 Z009 $4,043,546 $4,050,206 $4,055,474 $4,060,742 2010 $4,150,940 $4,157,600 $4,162,868 $4,168,136 2011 $4,258,335 $4,264,994 $4,270,262 $4,275,530 2012 $4,365,730 $4,372,388 $4,377,656 $4,382,924 2013 $4,473,125 $4,479,782 $4,485,050 $4,490,318 2014 $4,580,512 $4,587,176 $4,592,444 $4,597,712 2015 $4,598,162 $4,606,002 $4,612,198 $4,618,394 T otal $78,044 ,422 $7 8,176,452 $78,280 ,868 $78,385,284 116

PAGE 122

DETERMINATION OF COSTS All cost compo n ents are identified and discussed in the following paragraphs including lane construction, lane equipment, computer hardware and software, transponders/ decals, operating and maintenance, and contingency. Lane Constl'llCtion As discussed previously, a conservative estimate of lane construction costs is approximately $500,000 per lane. This estimate i n cludes planning, concept design, actual lane construction, right-of-way, and other associated construction costs. It is recognized that a per lane construction cost will vary significantly by location and region. For that reason, this conservative estimate of lane construction cost will be used which, if anything, overestimates the cost of lane construction. Lane Equipment The tables previously presented in Section VI represented low high, and mid-point cost estimates for the equipment required in various lane types The range of prices for lane equipment is based on the range of vendor offerings that have been observed in recent bids across the country. To some extent, the price ranges are reflective of line item quality and reliability. The lane types a long with the associated mid-point cost esti mates are summarized in Table 26. The mid-point cost estimates were used in the cost-effectiveness analysis. In addition, the cost of a booth must be included in all manual l anes including mixed lanes that include both A VI and manual. A conservative estimate for the cost of a booth is approximately $30,000. A breakdown of the number and configuration of lanes assumed under each alternative is provided in Table 27. 117

PAGE 123

TABLE26 Lane Equipment Cost by Lane Type l.aDO'I'ype Cost manual s 39,800 automatic s 90 ,750 A VI dedicated s 33,975 mixed (A VI and manual) $ 56,050 mixed (A VI and automatic) SlOS,SSO TABLE27 Number and Configuration of Lanes Under the Improvement Alternatives Year CoastnKIIoa 10% AVI 30%AVI 50% AVI 1990 8 automatic 8 automatic 8 automatic 8 automatic 4 manual 4 manual 4 manual 4 manual 1994 12 automatic 6 automatic 2 automatic 6 automatic/ A VI 8 manual 6 manual 6 manual 4 manual/ A VI 4 automatic/ A VI 6 automatic/ A Vl 2005 14 automatic 6 automatic 4 automatic 6 automatic/ A VI 8 manual 6 manual 6 manual 4 manual/ A Vl 6 automatic/ A VI 6 automatic/A VI 2 dedicated A VI 2015 16 automatic 8 automatic 2 automatic 6 automatic/ A VI 10 mauual 8 manual 6 manual 6 manual/ A VI 6 automatic/ A Vl 1 0 automatic/ A VI 2 dedicated A VI Computer Hardware and Sollware The computer hardware cost is also summarized in Section V1 of the repon and indicates a low, high, and mid-point cost estimate for both plaza hardware and host hardware. Just 118

PAGE 124

as with Jane equipment, mid-point cost is used in this analysis and represents a conservative cost estimate. This is particularly true when considering that the total cost of the host configuration is placed on this single plaza. If an A VI system were ultimately implemented, the cost of the host configuration would be spread across all plazas participating in the program. The mid-point cost estimate is $111,425 for the plaza configuration and $324,730 for the host configuration. These conservative estimates are encumbered in the first year of the analysis and are assumed to cover any necessary enhancements over the time period of the analysis. The resu lting total hardware cost is $436,155. Historically, the general range for plaza level software has been $150,000 for one or rwo plazas. However, if a larger system purchases the software (10 to 40 plazas), the price per plaza drops to approximately $50,000. This analysis uses the $150,000 estimate. In addition, host software ranges from $250,000 to $750,000 for highly customized, new application development software. J ust as with the host hardware configuration, the total cost of the host software is placed with the plaza under consideration. The mid-point cost of $500,000 for host software was determined to be a conservative estimate for the analysis. Following the conservative approach, computer hardware and software costs are attributed to A VI in each of the A VI alternatives. None of these costs are included in the full construction alternative even though, in reality, a portion of these costs should be attributed to both. Transponders/Decals The cost of a typical radio -f requency transponder ranges from $25 to $42, with a mid-point cost estimate of approximately $34. In addition, the cost of the typical bar code decal ranges from $1.00 to $2.00 depending upon the quantity purchased The resulting mid-point cost for a bar code decal is $1.50. To be consistent with the conservative estimates of other costs in the analysis, the mid-point cost of the radio frequency transponder is used since it is expected to be near the maximum possible unit cost for a t ransponder I decal. The number of transponders/decals that are purchased increases with the A VI participation rate. With a 10 percent level of participation in A VI, 10,600 transponders/decals were assumed to be purchased ( 10 percent of estimated total daily travel in 2015). With a 30 percent level of participation in A VI, 31,800 transponders/decals were assumed to be 119

PAGE 125

purchased (30 percent of estimated total daily travel in 2015). With a 50 percent level of participation in A VI, 53,000 transponders/decals were assumed to be purchased (50 percent of estimated total daily travel in 2015). This results in a costs of $360,400, $1,081,200, and $1,802,000 for the respective A VI improvement alternatives. All transponder/decal costs are encumbered in the first year of the analysis Oneratlne and Maintenance Probably the most difficult cost to identify and conservatively estimate is the operating and maintenance cost for Tarniarni Plaza The Turnpike was unable to provide actual operating and maintenance costs specifically for Tarniarni Plaza However, we were able to obtain fiscal year 1991 budgeted operating and maintenance costs by plaza for the Orlando Orange County Expressway The Orlando system is comprised of seven mainline plazas and numerous entrance and exit ramps, each of which are associated with a mainline plaza. The mainline barriers along with associated ramps include 115 lanes, 72 manual and 43 automatic. The total budgeted operating and maintenance expense for fiscal year 1991 is $9,310,392 If this cost were spread evenly over the 115 lanes, the resulting per lane cost is $80,960. As a result, the cost-effectiveness analysis assumes an $81,000 per lane operating and maintenance cost over the life of the analysis. This estimate is conservative in the sense that the ratio of automatic to manual lanes for the Orlando system is approximately 1:3 ( 1 automatic for every 3 manual lanes) while this same ratio is 2:1 (2 automatic for every 1 manual lane) for Tamiarni Plaza Therefore, it is believed that, since manual lanes are much more expensive to operate and maintain than automatic lanes, the per lane cost for Tamiarni Plaza is most likely much less than that assumed for the analysis. For this reason, $81,000 is deemed a conservative and appropriate estimate of per lane operating and maintenance cost for Tamiarni Plaza. Contingency Despit e the conservative cost and benefit estimates, a contingency cost of 5 percent was included to account for possible underestimations and for marketing and public relations costs. 120

PAGE 126

CONCLUSIONS As discussed previously, the indicators used to determine the economic desirability of the improvement alternatives include: Net Present Value (NPV) Benefit/Cost (B/C) ratio Internal Rate of Return ( IRR) As stated previously, economic desirability of an improvement alternative is indicated by an NPV greater than zero, a B/C ratio greater than 1.0, and an IRR greater than the assumed discount rate Table 28 summarizes the desirability indicators for each of the improvement alternatives in tile final year of the analysis. The full construction alternative has a net present value of $20,000,725, a benefit/ cost ratio of 1 .65 and an internal rate of r eturn of 29.8 percent, indica ting that this alternative is economically desirable. The implementation of A VI along with some plaza expansion also results in positive net present values, be ne fit/ cost ratios greater than 1, and internal rates of return greater than the four percent discount rate. This is true regardless of the A VI participation rate However, the economic desirability increase s dramatically when moving from the full construction alternative to 10, 30, and 50 percent A VI participation. TABLE28 Desirability Indicators Under Four Improvement Alternatives Coaslruclioo lO'li>AVI 30%AVI 50% AVI NPV $20,000,725 $25,760,623 $28,696.883 $34,420,348 B/C ratio 1.65 2.03 2 .29 3.07 IRR 29.80% 49 .00% 65.00% 103.00% Tables 29 through 32 provide the detailed results of the computations of these indicators for each of the improvement alternatives: full construction/ expansion, some construction 'vith A VI (10 perce nt participation rate), some construction with A VI (30 percent 121

PAGE 127

participation rate), some construction with A VI (50 percent participation rate). Each table provides the year-by-year benefits and costs and a cumulative computation of NPV and B/C ratios using a four percent discount rate. The results of the cost-effectiveness analysis clearly indicate that the economic benefits derived from the implementation of A VI will more than offset the costs of implementing such a program at Tamiami Plaza. The conservative methodology used throughout the analysis further supports this conclusion. 122

PAGE 128

Comp<)llnd lnkres t Fac tor U"r Y ear (PV) Benefits 1994 1.0000 $2,410,974 1995 0 .9615 $2,519,213 1996 0.9246 $2,627,452 1997 0.8890 $2,735,691 1998 0.8548 $2,843,930 1999 0.821 9 $2,952,167 2000 0.7903 $3,060,406 2001 0.7599 $3,168,645 2002 0 .7307 $3,276,883 2003 0.7026 $3,3&5,122 2004 0.6756 $3,493,353 2005 0.6496 $3,613,966 2006 0.6246 $3,721,361 2007 0.6006 $3,828,756 2008 0.5775 S3,936,1S1 2009 0 .5553 $4,043,546 2010 0.5339 $4,150,940 2011 0 .5134 $4,258,335 2012 0 4936 $4,365,730 20 1 3 0 4746 $4, 473,125 2014 0.4564 $4,580,512 2015 0.4388 $4,598,162 INTERNAL UTE OF RETURN : -------Table 29 Tamiami Plaza Benellt/Co s t Analysis Alternative 1 : Con struction Inves tment Opentlnall< ResidUIIt Costs Mafotcnaace Value $4,874,31 0 $ 1,620 ,000 so so $1,620,000 so so $1,620,000 so so $1,620,000 so so $1,620,000 so so $1,620,000 so so $1,620,000 so so $ 1 ,620,000 so so $1,620,000 so so $1,620,000 so so $1,620,000 so $1,240,515 $1,782,000 so so $1,782,000 so so $1,782,000 so so $1,782,000 so so $ 1,782,000 so so $ 1,782,000 so so $1,782,000 so so $1,782,000 so so $1,782,000 so so $1,782,000 so $2,437,155 $2,106,000 $3,210,315 Net Pnseol Benefit/Cos t v ..... Ratio ($4, 083,336) 037 ($3,218, 101) 0.60 ($2,287,261) 0.76 ( $1,295,416) 0 88 ($249,196) 0.98 $845,148 1 .06 $1,984,122 1.13 $3,160,965 1.19 $4,371,633 1.25 $5,611,785 1.30 $6,871,355 1.35 S7,261,Sll 1.34 $8,472,830 1.37 $9,102,059 L41 $10,946,027 L44 $12,201,783 1.47 $13,466,580 1.50 $ 14,737,864 1.53 $16,013,266 1.56 $17,290,588 U9 $18,567,793 1.61 $20,000,725 1.65

PAGE 129

Compouod Jnlerest Factor User Year (PV) Beaents 1994 1.0000 $2,416,212 1995 0.9615 $2,524,450 1996 0.9246 $2,632,688 1997 0.8890 $2,740,926 1998 0.8548 $2,849,164 1999 0.8219 $2,957,402 2000 0.7903 $3,065,640 2001 0.7599 $3 ,113,878 2002 0.7307 $3,282,115 2003 0.7026 $3,390,353 2004 0 .6756 $3,498,591 2005 0.6496 $3,620,630 2006 0.6246 $3,728,024 2007 0.6006 $3,835,418 2008 o.sn5 $3,942,812 2009 0.5553 $4,050,206 2010 0.5339 $4,157,600 2011 0.5134 $4,264,994 201 2 0.4936 $4,372,388 201 3 0.4746 $4,479,782 2014 0.4564 $4,587,176 2015 0.4388 $4,606,002 INTERNAL RATE OF RE1\JRN = 49.00\li> --Table 30 Tamiaml Plaza Benefit/Cost Ana l ysis Alternative 2: 10% AVI Participation lnvestmeot Operating & Residual Costs Maioteaaoce Value $4,058,465 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $1,271,655 $1,458,000 $0 $0 $1,458,000 $0 $0 $1,458,000 $0 $0 $1,458,000 $0 $0 $1,458,000 $0 $0 $1,458,000 $0 $0 $1,458,000 $0 $0 $1,458,000 $0 $0 $1,458,000 $0 $0 $1,458,000 $0 $2,437,155 $1,782,000 $3,226,632 Net Present Beatiii/Cost Value Ratio ($2,938,253) 0 45 ($1,757,051) 0.73 ($521,207) 0.93 $763,327 1.09 $2,090,978 1 .21 $3,456,529 1.31 $4,855,101 1.40 $6,282,133 1 48 $7,733,369 1.55 $9,204,833 1.61 $10,692,825 1.67 $11,271,586 1.64 $12,689,436 1.68 $14,117,252 1.73 $15,552,169 tn $16,991,529 1.81 $18,432,867 1.84 $19,873, 903 1.88 $21,312,527 1.91 $22,746,793 1.95 $24,174,908 1.98 $25,760,623 2.03

PAGE 130

Comp o u n d l oteres t Fact o r User Ytar (PV) Bendlts 1994 1.0000 $2,420,352 1995 0.9615 $2,528,590 1996 0.9246 $2,636,828 1997 0.8890 $2,745,066 1998 0.8548 $2,853,304 1999 0.8219 $2,961,542 2000 0.7903 $3,069,780 2001 0.7599 $3,178,018 2002 0 .7307 $3,286,255 2003 0.7026 $3,394,493 2004 0.6756 $3,502,731 2005 0.6496 $3,625,898 2006 0.6246 $3,733,292 2007 0 .6006 $3,840,686 2008 0.5775 $3,948,080 2009 0.555 3 $4,055,474 2010 0.5339 $4,162,868 2011 0.5134 $4 ,270,262 2012 0.4936 $4 ,377, 656 201 3 0.4746 $4,485, 050 2014 0. 4564 $4,592,444 20!5 0 4388 $4,612,198 INTERNAL RATE OF RETURN = 65,00% -----T a b l e 31 T a miami Plaz a B e n e fit/Cost Ana l ys i s Alterna tiv e 3 : 30% A VI P a rtic ipat i o n In ves t men t O perating & Residu al Costs Maintenance Valu e $3,686,345 $ 1 134,000 so so $1,134,000 so $0 $1,134,000 so $0 $1,134,000 $0 $0 $1,134,000 $0 so $1,134,000 so $0 $1,134,000 so $0 $1,134,000 so $0 $1,134,000 so $0 $1,134,000 so $0 $ 1 ,134,000 so $1,240,575 $1,296,000 so so $ 1,296,000 so $0 $1,296,000 so so $1,296,000 so $0 $1,296,000 $0 so $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 $0 $0 $1,296,000 so $1,343,003 $1,458,000 $2,061,455 Net Pres eo t Benefit /Cos t Value Ratio ($2,399,993) 0.50 ( $1,059,041) 0 82 $330,408 1.05 $1,762,640 1.22 $3,232,308 1.36 $4,734,415 1.48 $6,264,289 1.58 $7,817,575 1.67 $9,390,207 1.75 $10,978,399 1.83 $12,578,629 1 .90 $13, 286 ,233 1.85 $14,808,558 1.90 $16,336,831 1.95 $17,868,341 1.99 $19,400,579 2.04 $20,931,223 2.08 $22,458,130 2.1 2 $23,979,322 2.16 $25,492,980 2 19 $26,997,434 2 23 $28,696,883 2 29

PAGE 131

Compound Inte rest actor User Year (PV) Benefits 1994 1.0000 52,424,492 1995 0.9615 52,532,730 1996 0.9246 $2,640,968 1997 0.8890 $2,749,206 1998 0.8548 $2,857,444 1999 0.8219 $2,965,682 2000 0.7903 $3,073,920 2001 0.7599 $3,182,158 2002 0.7307 $3,290,395 2003 0.7026 $3,398,633 2004 0.6756 $3,506,871 2005 0.6496 $3,631,166 2006 0.6246 $3,738,560 2007 0.6006 $3,845,954 2008 05715 $3,953,348 2009 0.5553 $4,060,742 2010 05339 $4,168,136 2011 05134 $4,275,530 2012 0 .4936 $4,382,924 2013 0 .4746 $4,490,318 2014 0.4564 $4,597,712 2015 0.4388 $4,618,394 INTERNAL RATE OF RETURN = 103.00% Table 32 Tamiami Plaza Benefit/Cost Analysis Alternative 4: SO% AVII'articipation lovestmeat Operating & Residual Costs Malotenaoce Value $3,389,300 $810,000 so so $810,000 so so $810,000 so so $810,000 so so $810,000 $0 $0 $810,000 $0 $0 $810,000 $0 $0 $810,000 $0 $0 $810,000 $0 $0 $810, 000 $0 $0 $810, 000 so $71,348 $972,000 $0 $0 $972,000 $0 $0 $972,000 $0 $0 $972,000 so $0 $972,000 so $0 $972,000 so $0 $972,000 so $0 $972,000 so $0 $972,000 so $0 $972,000 so $1,230,705 $1,134,000 $1,329,698 Net Present Benefit/Coot Value Ratio ($1,714,808) 0.58 ($118,336) 0 .98 $1,574,497 1.27 $3,298,444 151 $5,048,601 1.71 $6,820,421 1.87 $8,609,629 2.02 $10,412,274 2 .15 $12,224,675 2.27 $14,043,414 2.37 $15,865,324 2.47 $17,546,321 253 $19,274,307 2.60 $21,000,329 2.66 $22,721,983 2.72 $24,437,052 2.78 $26,143,495 2.83 $27,839,439 2.89 $29,523,167 2.94 $31,193,110 2.99 $32,847,838 3.03 $34,420,348 3.07

PAGE 132

VIII. CONCLUSIONS AND RECOMMENDATIONS Several conclusions and recommendations can be made as a result of the evaluations and findings contained in this report. These recommendations are listed below. 1. Based on the favorable results of the conservative cost-effectiveness analysis, the Florida Turnpike user /focus group surveys, and the findings of this report in general, it is recommended that A VI be implemented on Florida's Turnpike. 2. Based on the travel patterns of existing commuters and the anticipated location and magnitude of Turnpike traffic growth, it is recommended that A VI be implemented in a three-step process. Figure 26 Ulustrates the recommended A VI staging process. The (lJ'st priority segment for implementation is from (and including) Lantana (Palm Beach County) to the southern terminus of the Turnpike, including the HEIT. This segment coincides with the first stage of the toll conversion program from ticket to coin and includes the area of the Turnpike where 185,000 (85 percent) of Turnpike daily commuters travel. Implementation phasing for t h e first priority segment cannot be specified at this time without further site-specific evaluations. However, implementation phasing should generally occur from the south to the north. The second priority segment for A VI implementation is from (and including) the Orlando West interchange south to the future Poinciana interchange, including the Bee Line West Expressway. This area coincides (in part) with the second stage of the toll conversion program from a ticket system to a coin system and would provide !QgU commuters the ability to "by pass" seasonal traffic (vacationers). The A VI system on this segment should be limited to local commuters until such time as the third priority segment is implemented. The third priority segment for A VI implementation includes all those remaining Turnpike sections not previously indicated in the first or second priority segments. The implementation of A VI on this segment is intende d for continuity purposes and long distance commercial and commuter traffic. In other words, only minimal A VI 127

PAGE 133

(julf of Ut>(jco BEE liNE WEST EXPRESSWAY ORLANDO FUTURE POINCIANNA INTERCHANGE FIGURE 2 6 RECOMMENDATIONS FOR AVI IMPLEMENTATION 1St PRIORITY All Mainline Barriers & Exll P lazas South ol Lantana Rd. (Malnline/H.E.F.T. Comm u t e r s ) 2nd PRIORITY Exit P lazas Be!Wee
PAGE 134

implementation (Le., one lane in each direction on mainline barriers only) is necessary. The implementation of A VI along this segment should coincide with the conversion of the ticket system to the coin system. If this conversion does not occur along the central portion of the Thrnpike, the provisions for continuity will have to be determined during the concept design phase. 3 The focus group representing potential commercial users of A VI expressed a strong desire to use an A VI system. The group cited the ability to track vehicles as being the major advantage of A VI, which in turn would also significantly improve their accounting procedures. For this reason, it is recommended that the A VI system implemented on Florida's Turnpike be designed to accommodate commercial applications. In particular, the configuration oflhe system must be such that single unit trucks and semi-trailer trucks can also use A VI. 4. The advantages and disadvantages of the three major ownership/fmance options were identified in this memorandum. Each of the ownership/finance options was detennined to be a reasonable alternative with the ultimate selection dependent upon the relative weights placed on the advantages and disadvantages by Turnpike officials. It is recommended that the Identified advantages and disadvantages undergo further review along with the specific objectives, policies, and legal constraints associated with ownership/finance options. 5. The evaluation of A VI technologies resulted in the identification of numerous issues including resistance to duplication (security), reliability, potential for "multiple reads" (speed vs. reliability), resistance to electrical interference (lane-to-lane), tolerance to environment, simplicity of tag (cost), and health/safety. Based on the relative comparative evaluation of technologies with regard to these issues, it is recommended that, since all four A VI technologies (i.e., RF/Microwave, SAW, Inductive Loop, and Bar Code) currently have successful applications in operation, all are suitable Cor implementation. However, it is also recommended that Turnpike officials specifically establish the importance of the identified issues in order to select the technology which best fits Florida's Turnpike. 129

PAGE 135

It is also recommended that the Turnpike work closely with Orlando Orange County Expressway Authority since they are currently in the process or acquiring and perfonning acceptance testing ror an A VI system. It would be particularly advantageous to work together with the intention of ultimately implementing the same technology. An obvious objective would be to have a standard A VI system throughout the state. 6. Research into the issue of dedicated versus mixed-use A VI toll Janes bas resulted in the establishment of general thresholds (based on the magnitude of traffic flow) for when to consider various A VI configurations. Mixed-use A VI should be considered when experiencing 3,000 vehicles per hour; dedicated A VI lanes should be considered when experiencing 5,000 vehicles per hour; and express A VI lanes should be considered when experiencing 7,000 vehicles per hour Each of these guidelines assumes that this traffic flow i s for the peak-hour, peak-direction. It is important to recognize that these guidelines are based on ideal plaza Jane configurations with reasonable levels of A VI participation (at least 20 percent). The selection of dedicated or express A VI lanes may also be an integral part of marketing the system to the Turnpike patrons. H this is the case, then various A VI configurations should be considered much sooner in the process. As a result, it is recommended that the issue of dedicated vs. mixed-use A VI lanes be considered on a site-by-site basis through the use or the identified guidelines and the proposed marketing program. 7. Based on the greater average toll processing speeds associated with A VI, it is recommended that the current Florida Turnpike procedures regarding tramc control (pavement marldngs, signage, channelization) and tramc rules (speed limits and passing areas) be re-evaluated and revised as necessary to safely and emciently accommodate (and direct) the use of A VI. This will help to ensure the safety of all Turnpike patrons as well as toll collectors. 8. Current Florida statutes do not pennit enforcement of toll violations by photographing a vehicle and citing the vehic l e owner through the mail As it stands, toll violators must be physically stopped at the location where the violation occurs in order for the violator to be penalized. As a resu l t, it is recommended that the Turnpike propose legislation to revise the Jaw in order to permit photographic 130

PAGE 136

enforcement and the ability to fine the owner of a vehicle (similar to a parking ticket) as opposed to the driver of the vehicle. 9. It is believed that the major reason for the use of gates on the Turnpike is for a means of deterring violations. However, the focus group surveys indicated a strong negative response to any barriers in the lane and that the level of participation in A VI would likely increase if gates were eliminated. For these reasons, It is recommended that, if legislation can be passed which legall2es photographic enforcement, then the use of gates should be eliminated. However, if photographic enforcement is not pennissible, then it is recommended that the Turnpike continue to use gates as a method of enforcement. 10 Based on the Turnpike policy which prohibits the extension of credit and the need for the Turnpike to avoid additional costs and time associated with collecting on delinquent accounts receivable, it Is recommended that a post-payment option should not be offered to A VI patrons. However, it is recommended that various pre. payment options be offered including cash, check, credit card, and electronic fund transfers. II. Based on the evaluation of discount, premium, and normal toll structures, it Is recommended that A VI be implemented at the nonnal toll rates and that premium toll rates should not be utilized. Finally, it is recognized that discount tolls may be an effective method for marketing an A VI system. There is some uncertainty as to whether offering discounts would be legal based on the current Florida Turnpike Bond Official Statement. As a result, it is recommended that further evaluation be conducted by the Turnpike legal counsel into the legality of offering discounted tolls to A VI patrons if at some point this payment option is deemed appropriate ror marketing purposes. 1 2. Based on the conclusions of the commuter and commercial user focus groups, it is recommended that the transponder be provided to patrons under a lease agreement requiring a modest payment of $2 to $3 per month. In addition, it is recommended that an optional monthly account statement be offered at a cost of approximately $2 per month. 131

PAGE 137

13. The level of participation in an A VI system is extremely difficult to estimate. The mail-back survey indicated nearly 70 percent of the 2,088 respondents would be interested in using A VI if the toU rates remained the same. Other systems around the country have experienced a range of participation rates from 20 percent in Dallas to approximately 60 percent in Treasure Island, Florida. It is reasonable to expect that an e.fficiently operated A VI system on Florida's Turnpike would realize a participation rate of around ZO percent. However, it is recommended that a detailed marketing program be established as part of the A VI bid specification. This marketing program could include distributing infonnatlon in order to educate the public about AVI, patron service cellters (as ill Dallas), and other methods of marketing a product or service. An effective marketing program could be expected to have a significant impact on the level of participation. 14. Finally, based on the findings of this report and the unknown site-specific conditions regarding A VI implementation phasing, it is recommended that the next phase or A VI implementation be a concept design phase. This phase would basically consist of a six to twelve-month effort to detail and evaluate the specific needs of the Turnpike with respect to an A VI system. This would include a more-detailed consideration of the phasing of A VI implementation. a more-detailed estimate of the costs of A VI implementation on a site-by-site basis, technology acceptance testing, and the preparation of an A VI bid specification. An estimate of this phase of A VI implementation is approximately $450,000. CUTR would like to continue to play a role in the coordination and review of the next phase. As part of this coordination effort, CUTR would assist the Turnpike In the selection of a systems integration consultant who will help to expedite the efforts of the concept design phase. Finally, CUTR believes, if this approach is followed, an A VI system could be operating on Florida's Turnpike by July 1, 1993. 132

PAGE 138

APPENDIX Florida Thmpike Oral In terview Surv e y Form Florida Turnpike Mail Back Survey Form Participants in Commercial User Focus Group D-QUEUE Model Output -1990 Tamiami Plaza D-QUEUE Model Output 2015 Tamiami P l aza D-QUEUE T oll Plaza Simulation Runs 133

PAGE 139

94.9\ 61.7\ 18,0\ l.4\ !.7\ 1 7.3\ 4,0\ 6.2\ 6.5, Aorida Turnpike Oral Interview Survey Fonn 1. Vehicl e Type a passenger car Q single trUd< semHraiier 0 other 2 Veh icle Oc:c:upancy o one c two c three o IT'tCtoretc!e 2' o bus .2\ o recreational veh.iC:e 3, a four 2.0' a ftveormore 1 .2\ Good Moming/Gllod .Allemoon. Aorida's Turnpike iS conduc:ting a survey that will help to Improve thiS facility. I would like to ask you three quesdons. 3. 4. s. Are you a Aorida resident? a Yes o No W hat iS the purpose of your 1rip? a Wori< Comrriute o Company Business o Medical/Dental o Vacalian a Sdloo l 2.2 c Shopping 1 s. a Social 6 a Other(peesespeciiy)_.::: 2:.:.;... _ How many days per week do you 1ravel on this !acillty7 (c:11ec:k only one) o < 1 day/week o o 1 dayfweek o o 2 daysfweek o > s days/week o 3 days/wee k Thank you t a r your cooperation i n thiS interview. Please fill out this mail-back survey 1onn as soon as possible to compl ete tne seconct part ot tnls survey No postage iS required. Thank you, again. 134

PAGE 140

:..!.l t.oc&'l:ions C:C.:::incQ (N,08.81 F",.auCA TURHPIKC MAil..aACX SURVEY FORM .!icrid'3 T'umpike is condudnq 1 swwy tftlt vrilf lmp'oYe U'UI tactnry. 1hant yau for j)lrttc::pa:Wtq i n Ute roadsloe i nterView oortlcn Of tne surv.y. P*-lUI Ol41ftta llJIVeY form aoonaa poaaiblil to eomcifte U'\e sec::n a i'l&lf of Uta survey. Ho 9C*R'Q8 ia Yow ia 7urncike me pocxandl,l tel used .in aiJICmlliC whic: (AV1) sysem u .a mC".o: of :ell :::IK:"".on. The AVI systlfTI 'W!Culd ryoically with .t C"8dlr: c::uu...a tcil I3Q 1l"'lll is mCW'Qd on ::"le Y&niC:t. ;.r. reaa U'le tOll wnen it I)U&U U'lrCuQn a. ted cdlec:scn :an.. 7"r.:is O'VISIClCn iS 7!.., .autcmotr....uly ::: an numow W\iout to trial veniCe. The l:s a rcll wnicl'\ tr.e neeo t or :3Sll. !lCl
PAGE 141

PARTICIPANTS IN COMMERCIAL USER FOCUS GROUP Baron Messenger Service, Inc. Co nsolidated Oil Co. Florida Power & Light Florida Truckers Association Publix Rechtien International Southern Wine & Spirits Thomas W Ruff & Co. Wingerter Labs 136

PAGE 142

IHIIEl VI. lOOt A ilyoMic Toll PlZi a.-;ng Analysis ProtJr By: FIOO"ido Stott Deportof Trans.,...tation FellNory1 l'l'JO Tnffic Engi.-ing Tel: locatiotu 1990 T.-iai M&il"' LiM B&rr-ier Cotdition:Ptill Hoctr 0.111 08-27-l'l'JO Toll Sttion Sorvict R.alest Sh. II: 350 11/th/Hrl st . te: 350
PAGE 143

IHl.ELE Yl.IOO: A Toll Plu Aoalyois Proqru By: Flcrid Stilt of TronslX>rbtim Ftlruuy,1990 Tr1ffie Enginooring 2015 Twui lllin Line llin"ior c..-c!ition:Ptil< lblr 01111 08-27 Toll Sl1lion S...iet Ritts: S\1. Ill J:iO ll'oii/Hrl 511. 12: J:iO 1\ltii/Hr) 511.13: 500 (Yefl/lr) st .. M: 500 IYohtHrl st .. soo ll'ollllrl S\1.16: 500 IVellllrl st .. t7> 500 ll'ollnrl No. of Hi .... y Lints: J SIR11tim SEED: 3571 07 00 07 IS 07 30 07 45 1M. 'ISIS IIERIOD 0715 0730 0745 0800 YDL.LIE US NINI 738 8J4 867 171 ARJIII'Cl. 1\QI 715 806 &71 701 SERVICED 1\QI 706 m 801 7'39 IIAX QSE 1\QI J 1% 14 AYE UJ 2 4 9 10 AYE IIEl.AY I I 12 2S 72 138

PAGE 144

4 Booth 4 Boo th.s 4 Booth.t 3 Booth$ 2 Booths 4 Booths 5 Booths DWEUE TOll SIMULATOR Colwentionat Set-Up Manned Lane 350 VPH Automatic Lane 500 VPH Mh:ed/AVI : 700 VPH Exclusive AYI a 1200 VPM separated AVJ 1800 VPH 1000 VPH (230,250,280,240) 15 Min. Increments 2 Lane Approach 3 Memed 1 .t.utc.ttc 2 lane A,pproech Z Mamed 2 Autc.tic 2 Lane Approach 2 Memed 1 Nixed AUtc:.etlc 2 Lane Approaeh 2 M tnned Auto.tic Z Lane Approoch 1 Memed 1 2000 VPH (460,500,560,480) 15 Min. lncr...nta 2 Lane Approach 2 Mamed 2 Lane Approach 3 Manned 2 2 Autc:.etic ..... ..... .... ..... --.... .... Q= 3,3,3,3 G 3,3,3, 3 Q 3,3,3,3 Qz 3,3,4,4 Q 5,20,48 STOP Q 6,22,54 S TOP Q 3,5, 16,16 4 Booths 2 Lane Ar:9f"08Ch 1 Menr'led 3 --Q= 4,10,35 STOP 5 Booths 2 Lane Approach 2 Memed Max. 0= 3,4,7,6 I I I 3000 VPH (690,750,840,720) 15 Min. lnc.--nts 1 s Booths 1 6 Bootha 1 1 Booth 1 r Bootht 1 1 Bootht 2 Lane Approtch 2 Lane Approach 2 lane Approach 2 Lane 2 lanl Apt:H"OICh 1 Mamed 1 MaM 15 Min. Increments 8 Booth-s 2 L.,. A 8 Booths 2 Lane Approach 3 Memed 2 Mamed 4 AUto.tlc 5 4000 VPH Prorated at (7/9) (716,778,871,747) 15 Min. Incrementa 9 Booths 2 Lane Approach 9 Booths 2 l.ane Approach 3 Mamed Z Mamed 4 5 Autc.etic 4000 VPM Prorated at C7/10> (644,700,784,672> 15 Min. Jne remonts 10 Soot h s 2 Une A pproach 4 Mamed 3 AUtOMatic Max. 0= 2'0 STOP Max. M ax. Qa 3 ,3,9.9 Qa 3 ,5, 16,16 Max. Q 12,12.2.6 Mtx. Q 9,28,72 STOP Max. Qz 8,31 STOP Max. Qz 7,25,29,39 Max. Qz 4, 1 5 16,1 8 Max. Q: 3,8.24.24 Max. Q:z:: 3, 14,13,1 3 Max. Oo 3,5,12,12 I tnc:Ucates Acceptable Toll lane Configuration 139

PAGE 145

DOUEUE TOU SUIUL.ATOR RUN S Conventional S e t -Up 1---4664 VPN Prorated at (7/13) (577,680, 761,653> 15 Min. lncre.ents at T .. iami 13 Booths 2 L ane Approach 13 Booths 2 Lane Approee h 13 Booths 2 Lane Approach 4 M ennod 3 Momod 2 Mamed 3 AUt<*tiC 4 5 Aut0118t;c 5000 VPH Prorated a t <7Jfn (895,973,1089,934) 15 Min. Incr ements 9 Booths 9 Booths 1 0 Booths 1 0 Booths 11 Booths 11 Booths 11 Booths 3 lane A pproach 3 L ane Approach 310ennod 2 llanned 4 AUtONtlc 5 A utOfl'llt I c 5000 VPH Prorated a t ( 7 / 10) (805,875,980,840) 1 5 M i n lncrements 3 Lane Approach 3 Momod 4 AUtQIIIIt I e 3 Lane 2 Momod 5 AutCIIIItt i e 5000 VPH PI"''f"a ted at (7/11) (731,795,890,763) 15 Mfn. lnc:r...nts ----3 L ane Approach 4 Named 3 3 Lane Approach 2 M on>ed 5 AutONtie 3 Lane Approae h 3 Memed 4 Autc..tio 500 0 VPH Prorated a t (7/12) ( 6 7 0,729,816,700) 15 Min. Increments 1 2 Booths 3 lane Approach 1 2 Boot h s 3 LaM 4 Mamed 2 Memed 3 Memed 3 A utomatic 5 AutOIII&t; c 4 Automat;c 1 2 Boot h s 3 Lane Approach NIX. o3,4,10, 1 0 MIX. Qa 3,3,6,6 Max. Q-= 3,3,4, 3 M ax. Q= 16,45,96 S TOP M a x Qz 1 1,41 STOP .... Qa 6,20,48,60 """ Q 5,13,40,45 ... Q:l 5,13,35,39 .... Q z 3,8,20,20 .... Q:o 3,10,29,32 M I X Qa 3,5,15, 16 M a x Q 3 ,4,9,8 M ax Q 3,4 11,1 1 1-----------6000 VPH at (7/12) (805,875,97'9,840) 1 5 M fn. Increments 12 Booths 4 Lane AA>roach 12 Booths 4 Lane App rotch 1 2 Booths 4 Lane Appr oach 2 Msmed 4 Memed 3 Mamed 5 AutOIMtie 3 AutOIIIIIt i C 4 600 0 VPH P rora t ed a t (7/13) (742,807,904,775) 1S Min. Increments 13 SOOths 3 Lane A pproach 13 Booths 3 Lent Approach 13 Booths 3 Lane A pp ro.c:h 4 Mon>ed 2 Mamed 3 Mennod 3 5 Auto.atic 4 AUtOIMtic 6000 VPH Proreted at (7/14) (690,750, 840 ,720) 15 Min Inc r ements --------1 4 Boot h s 3 Lan. Approach 14 Booths 3 Lane Appf"oaeh 1 4 Booths 3 Lane Approach 4 .. mod 2 ........ 3 Msmed 3 Aut0fl'l8t; c 5 AutOIIIIit f c 4 Automatic M&K. QD 4,14,43,52 Max. G 6,29,67,85 M.8J(. G 7,18,49 STOP MctlC. Qa 5,19,44 STOP MaliC. G= 3,7,28,28 Max. a-4, 1 2 ,32,32 M ax. Q 3,8,26,27 Max. O 3,4, 11,11 Q : 3,7,15,16 1 -----------------140

PAGE 146

OOUEUE TOLL SUIJLATOR$ Conventionet Set-Up 7000 VPH Prorated at (7/15) (752,817,915,785) 15 Mfn. Increments --------------------------------------------15 Booths 3 Lane Approach 15 Booths 3 Lane Approaeh 15 Booths 3 L ane Appt"Oaeh 4 Mamed 2 Mamed 3Momed 3 AUtOMtic 5 Autc-etic 4 Autc-etic 1000 VPM Prorated at (7/16) (705,767,858,136) 15 Mfn. Increments ---------------------------------------------16 Booths 3 lane 16 Booth.$ 3 lant 16 Sooths' l Lane Approach 410annod 2118Med 3 .. mod 3 AutDMtic 5 Aut01118t i c 4 AutOfllet i c M I X. Qa 6,17,42,46 Max. G 3,7,23,24 Max. Q 4, 13,37,39 Nax. Qa 3,11,32,33 Max. Qa 3,5,16,15 Max. 3,7,22,23 ---------------------------------------------------------------------------------------------141

PAGE 147

DQUEUE TOLL SIMULATOR RUMS Conventional and AVl SetUp t---------------------------------------------------------------------------------------ZOOO VPH (460,500,560,480) 15 Min. Increments -----------------------------------------4 Booths 2 lane Approeeh 2 MIMed 2 Autc.tic MIX .. 6,ZZ,54 STOP S Booths 2 l int Approld'l 3 2 MIX .. 3.5, 16,16 S Booths 2 lane Approaeh 2 Manned 2 1 ""'-' MaA .. 3,3,4,4 5 8ooths 2 lane Approach 3 Mamod 1 Automatic 1 Mixed Max. 3,4,7,6 4 Booths 2 Lane Approach Z MOMed 1 Autometic 1 Mixed Max. 4,11,24,28 4 Booths 2 laM Appro&eh Z"*'"
PAGE 148

OQUEUE TOLL SII JLATOit RUS Conventional and AVl Set-Up ------------------------------------------------------------------------------5000 VPH (1150,1250, 1 400,1200) 15 Min. 1 Booth s 2 Line "-""" 4 M anned 3 Auto.etie Max. Qa > > 15 veh. T Booths 2 Lane Approach 7 Mixed MaX. Q 4,8,22,2 1 6 Booths 2 lent 5 Mixed t Excluaiw Max. 0. 6 20,42 STOP 6 lootha 2 Lane Approach 4 Ntxed 2 exclusive Max Qa 5,5,7, 6 5 B ooths 2 lane Aj:JH'DKh 3 M ixed 2 Exclusive M I X Qa 8,32 STOP 7 Bootha 2 Lane Approac h 6 ltixed 1 E xclusive M ax o-4,4,5,4 600 0 YPH (1380.15 00,1 680,1440) 15 Min. Jner.-ents 7 Booths 2 L ane Approach 6 Mixed I .... Qc 5,5,3,4 6 Booths 2 Lane A pproach 4 Nixtd 2 Exclusive """ Q 12,11,3,5 6 loothl 3 Lane A pproach 5 Mixed 1 E.Jcclua ive """ Q 34,87 STOP 1 Sooths 4 Lane Awoach 6 Mixed 1 Exclusive .... Q 1 1,31 STOP 6000 VPM Prorated at (7/8) ( 1207, 13 13, 1470.1260 ) 15 M i n Increments 8 B oo t h s 3 lane 7 Mixed M8X. G 4,17,17,.8 600 0 VPH Prorated at (7/9) (1074, 1 167,1307, 1120) 1 5 Min. Increments 9 Booths 4 len. Appr'06C:b 7 Mixed Max. o-4 5,13,.13 ------------------------------------7 Booths 7 Boot hS 7 Booths 7 SOOths 6 Booths 8 Boot h s 8 Boot h s 8 Boot h s 8 Boot h t 8 Booths 9 Booths 9 Booths 9 Booths 7000 VPM (161 0 1750, 1960,1680> 15 Min. lncr..,ts -------------4 Lane Approaeh 6 M i xed 1 Exclusive 4 Lane Appro6Ch 4 Mixed 3 Exc lusive 4 Lane 3 Mixed 4 Exclusive 4 Lane Approach 4 Mixed 2 Exclusi v e 1 Seper-attd 4 Lane Approach 4 Nbed 1 Exclus ;ve 1 Separa ted 7000 VPII Prorated a t (7/8) ( 1 409,1531,1715,1470) 15 Min. Increments ------------------------------------4 laM Approach 7 Mixed 4 Lane Approach 6 Mixed 1 Exclusiv. 4 L ant Approach 5 Mixed 2 Exclusiv. 4 Lane Approach 4 Mixed 3 Exclusive 4 Lane Approach 3 Mixed 4 Exelu:r.ive 7000 VP14 P rorated at (7/9) (1252,1361, 1524,1306) 15 Min. lncrellllef'ltS ---------------------------------4 Lane ApproKh 4 l ane Approae h 4 Lane AJ)proae h 7 Mixed 6 Mhed 1 exc:lusiw 5 Mixed 2 E xclusit ..... Q z 41,98 STOP ..... Qz 8,28,28 STOP Max. Qz 5,1 0 ,9,5 M IX.. Qs 5,7,6,6 Mx. Qa 3 1,80 STOP """ G 31 STOP .... Qs 14,40 STOP Max. Qz 6,13,44 STOP .... Qa 4,5,21,21 Mu. Qa 4,5,6,6 Max. 0=11 9,29 STOP Nax. a. 5,6,31.31 NIX. Qa 4.5,8,7 1 ---------------------------------------------143

xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200277Ia 4500
controlfield tag 001 023228643
005 20110909162351.0
008 920507s1990 flua 000|0 eng d
datafield ind1 8 ind2 024
subfield code a C01-00081
b A53 1990
0 245
Analysis of automatic vehicle identification technology and its potential application on the Florida turnpike.
n Technical memorandum 2 /
by the Center for Urban Transportation Research.
Tampa, Fla. :
University of South Florida, College of Engineering, Center for Urban Transportation Research,
iv, 143 leaves :
ill. ;
28 cm.
Prepared for the Florida Dept. of Transportation.
"October 1990."
Also issued online.
Toll roads
z Florida
x Automation.
Motor vehicles
Automatic location systems.
Florida Turnpike (Fla.)
1 710
Dept. of Transportation.
University of South Florida.
Center for Urban Transportation Research.
USF Faculty and University Publications.
t Center for Urban Transportation Research Publications [USF].
4 856