W. L. Eisele, Texas Transportation Institute, Texas A&M University System,3135 TAMU, College Station, TX 77843-3135. M. W. Burris, Department of Civil Engineering, Texas A&M University, 3136 TAMU, College Station, TX 77843-3136. H. T. Wilner, HDR, Inc., 770 E. Warm Springs Road, Suite 360,Las Vegas, NV 89119. M. J. Bolin, Texas Department of Transportation, ShermanArea Office, 3711 US 75 South, Sherman, TX 75091.

68

Transportation Research Record: Journal of the Transportation Research Board,No. 1960, Transportation Research Board of the National Academies, Washington,D.C., 2006, pp. 68–79.

The Texas Department of Transportation (TxDOT) sponsored researchto develop a decision support tool to aid in evaluating key issues relatedto adapting a high-occupancy vehicle (HOV) lane to a high-occupancy toll(HOT) lane. The tool includes three broad categories of factors to con-sider: facility considerations, performance considerations, and institu-tional considerations. Facility considerations, such as design, operations,and enforcement, which have been shown to be critical factors, can pre-sent insurmountable obstacles to the implementation of HOT lanes. Per-formance considerations and goals allow the user to estimate the likelylevels of usage and person movement, factors that always bear signif-icantly on HOT lane development decisions. Institutional considera-tions are also addressed, as factors of interagency cooperation and legallimitations are historically important for HOV lane and HOT lane deci-sions. Finally, the research incorporates simple trade-off tools to allowTxDOT and local entities to assemble all relevant factors into an analy-sis to aid decision makers in evaluating the available options. The analy-sis tool was developed in Visual Basic.net. The program is called theHigh-Occupancy Toll STrategic Analysis Rating Tool (HOT START),and it is designed to be tailored easily to local needs. An application ofthe tool to the I-10 (Katy Freeway) in Houston is also provided. Manystate departments of transportation are facing the difficult task of per-forming assessments of potentially adapting HOV lanes to HOT lanes,and this paper describes a practical tool that can assist such analysis.

High-occupancy toll (HOT) lanes offer drivers the option of travel-ing on a high-occupancy vehicle (HOV) lane for a toll when theywould not normally meet the occupancy requirements of the lane.These characteristics have led to the growing perception that HOTlanes offer both substantial revenue opportunities and a solution toconcern about underused HOV lanes.

As of March 2006, HOV lanes have been adapted to HOT lanesin only five projects. However, numerous cities are in various stagesof implementing a HOT lane, according to the Internet site valuepricing.org (1). Transportation departments and transit authoritiesare aware that there are complexities and costs associated with adapt-

ing HOV lanes to HOT lanes and operating HOT lanes, but the exactnature and magnitude of these issues are generally unknown.

The complexities and costs associated with adapting HOV lanesto HOT lanes necessitate detailed evaluations of such projects. Fur-ther, each project is case specific, and the importance or relevanceof the numerous factors that must be considered in adapting anHOV lane to a HOT lane vary from one project to the next. Thoughdetailed analysis of the factors is necessary before dedicating finan-cial resources to such a significant transportation improvement,there is a need for a sketch-planning tool that can evaluate the mul-tiple factors (quantitative and qualitative) involved in implementingan adaptation project.

This paper describes a research effort sponsored by the TexasDepartment of Transportation (TxDOT) to develop a much-neededsketch-planning tool for assessing HOV-lane to HOT-lane adaptationprojects—the next challenge in the evolution of HOV facilities. Iden-tifying key issues and incorporating them into an evaluation toolare anticipated to be of benefit to the myriad of regions consideringthe adaptation of an HOV lane to a HOT lane.

This research project evolved from more than two decades of ex-perience with HOV lanes in Texas. The Texas Transportation Insti-tute (TTI) has teamed with TxDOT and the transit authorities inHouston and Dallas to perform ongoing, comprehensive evalua-tions of existing and proposed HOV lanes and HOT lanes since1979. This research project captures the benefits of this extensiveexperience in a manner that is not only applicable to Texas projectsbut also readily applicable to HOV-lane to HOT-lane adaptationseverywhere.

PROJECT OBJECTIVES

The research team initially developed a list of the most likely goalsbehind the adaptation of an HOV lane to a HOT lane. These goalsincluded

• Increase corridor mobility,• Increase corridor safety,• Generate revenue,• Make environmental improvements, and• Provide travel options.

Researchers then developed a list of the primary measures ofeffectiveness (MOEs) of these goals and the issues and elements thatwould prevent obtaining each goal. These items were then groupedinto three main categories—facility considerations, performanceconsiderations, and institutional considerations—as follows:

Analytical Tool for Evaluating Adaptationof a High-Occupancy Vehicle Lane to a High-Occupancy Toll Lane

William L. Eisele, Mark W. Burris, Hannah T. Wilner, and Michael J. Bolin

1. Identify, analyze, and quantify the facility considerations in apotential adaptation of an HOV lane to a HOT lane. This objectiveincludes those design, operations, and enforcement features or char-acteristics that would be essential or desirable for a successful HOTlane operation.

2. Identify, analyze, and quantify the performance considera-tions associated with the adaptation of an HOV lane to a HOT lane.This objective includes how best to measure and predict the poten-tial for an adaptation project to accomplish the goals of the trans-portation agencies and communities involved in the project. Thesegoals might include increasing person-movement, reducing conges-tion, generating revenue, providing travel options, and achievingother performance goals.

3. Identify, analyze, and quantify the institutional considerationsin evaluating the appropriateness of adapting an HOV lane to a HOTlane. This objective includes factors such as public acceptance, revenue use, interagency cooperation, and media relations.

In addition to the three categories (and related objectives), it wasa project objective to develop an appropriate mechanism (analyti-cal tool) to allow public agencies to evaluate the tradeoffs withinand among the project objectives listed above. It is unlikely that anypotential HOV lane project represents an ideal combination of fea-tures, demands, and characteristics to ensure success as a HOT lane.Satisfying this objective allows the analyst to assess the relative sig-nificance of tradeoffs among facility, performance, and institutionalobjectives and considerations in reaching decisions about the mostappropriate decision. The result will be a user-friendly evaluationtool for considering the trade-offs and communicating the findingsat a sketch-planning level. Potentially viable projects should thenbe examined carefully for their net societal costs and benefits.

The remainder of this paper discusses key elements of theresearch that were performed to address each of these objectives.

Eisele, Burris, Wilner, and Bolin 69

ANALYSIS PROCEDURE

The researchers and TxDOT personnel developed a lengthy list of fac-tors that have been identified throughout the documented research ashaving had some demonstrated or suspected degree of impact on theHOV-lane to HOT-lane adaptation. That list was consolidated to thosefactors that could have a meaningful bearing on the decision to adapt.

Once the key factors were identified, described, and bounded, theresearch focused on how to incorporate these relevant factors intoan analysis of the whole set that was logical, comprehensive, andexplainable. That process took into account three dimensions foreach factor:

• Weight—how significant or important this factor is relative tothe goals of adaptation;

• Score—how well this factor compares with a desirable or min-imum standard; and

• Interaction—how this factor interacts with other factors andhow that can be captured quantitatively.

Each dimension required comprehensive development, which isdescribed in subsequent sections.

With the large number of factors and detailed guidance associatedwith each one, a hard-copy workbook was not practical, so the TTIteam developed a software tool that accomplishes two tasks:

1. It guides the user through logical steps in the development ofan assessment and

2. It performs all the recording keeping and calculations automatically.

The analytical process is illustrated in Figure 1. The analyst isassumed to be a staff person in a transportation organization who

PerformanceData

FacilityData

InstitutionalData

InstitutionalAnalysis

& Direction

STAFF ANALYSISDISTRICT ENGINEER

ASSESSMENT &DECISION

TechnicallyAdvisable?

PerformanceAnalysis

FacilityAnalysis

END

NO

YES

InstitutionalAnalysis

FIGURE 1 Decision flowchart for adaptation of HOV lanes to HOT lanes.

has access to routine design, operations, and performance informa-tion. By using that routine design, operations, and performance infor-mation, along with links to additional information embedded in thesoftware program, the analyst prepares the analysis of the facility andperformance categories and prepares the input data for the institu-tional category of factors. While the analyst may conduct part of theinstitutional analysis, the final elements are likely left to a seniormanagement individual, who may be more likely to appreciate thepolitical sensitivities and interagency cooperation issues. At TxDOT,this individual is assumed to be the district engineer, the rankingstaff person for a geographic region of several counties, though theduties certainly could be delegated.

70 Transportation Research Record 1960

The following discussion addresses the individual componentsof the analysis in more detail.

Definition of Factors by Category

There are numerous factors to consider when an adaptation from anHOV lane to a HOT lane is investigated. Researchers separated thesefactors into facility, performance, and institutional categories to meetthe project objectives specified previously. The many potential factorswere narrowed down to those anticipated as the most important in eachcategory. These factors are shown in Tables 1 through 3 for facility,

TABLE 1 Facility Factors When Considering HOV Lane to HOT Lane Adaptation

Factor Description and Questions Addressed Default Weight

Facility cross section

Lane separation for toll collection

Facility access satisfiesO-D requirements

Facility access design

Ability to enforce

Facility traffic control

Pricing strategy

Incident management

Maintenance

This factor is concerned with the design envelope available along the proposed HOT lane. The AmericanAssociation of State Highway and Transportation Officials (AASHTO) Guide for High-Occupancy VehicleFacilities (2) provides examples of cross sections for barrier- and buffer-separated HOV facilities. Thesecross sections, including lane width and shoulder width, are typically applicable to HOT facilities. Typicalquestions include

Is there adequate space to bypass a disabled vehicle?For buffer-separated facilities, is there adequate space for a vehicle to avoid an encroaching vehicle?

This factor is concerned with the adequacy of lane separation between HOT and GPLs to support tollingoperations. Three types of lane separation can be considered, each with advantages and drawbacks:

Rigid barrier—concrete barrier separates HOT lane from GPLs. Flexible barrier—also known as plastic channelizers, pylons, or candlesticks separating HOT from GPL.Buffer—striped separation, varies in width and may consist of “double double” lines or raised pavement makers.

The principal consideration for this factor is, “Do the access points serve potential HOT lane demand?”Answering this question begins with defining the primary or target users of the facility. HOV lanes aredesigned to serve buses, carpools, and long-distance commute trips. If the facility becomes a HOT lane, willthese still be the primary users? Do lower-occupant vehicles buying into the lane have a different set oforigin-destination (O-D) patterns? By defining the primary or target users, in priority order, along with theirO-D patterns, the location of access points can be determined based on how best to serve their needs.

The design of access points can impact the operation of both the HOT lane and adjacent GPLs. This factorevaluates the access design in terms of the ability to meet guidelines developed in Texas research (3) and othernationally accepted guidance. There are three types of access that can be provided:

At-grade slip rampDirect connect rampNo designated access (continuous)

HOT lane enforcement involves verifying occupancy requirements as well as toll account validity. This factorasks the question: “Can adequate compliance be achieved through planned enforcement operations?” Thereare three areas of consideration:

Adequate space for occupancy verificationEase of occupancy checkLevel of law enforcement

Signing, pavement marking, and other forms of driver communication can be challenging for HOT lanes forseveral reasons. First, the HOT lanes are located in an existing freeway corridor with their own set of signingneeds and requirements, sometimes conflicting with messages and information requirements for drivers in theHOT lanes. This creates the potential for confusion and information overload. Second, there are additionalmessages for a HOT lane operation that are not necessary for a typical HOV lane, namely price level that canvary by time of day and/or user group. This facility factor poses the question, “Can effective drivercommunication be accommodated when adapting to a HOT lane?” Since there is limited specific guidanceavailable in the Texas Manual on Uniform Traffic Control Devices, the general guidance and best practicescomes from Texas research (4) and current research with the Federal Highway Administration that describes

Defining target users and their information needs,Signing features, andSigning placement.

Pricing strategy refers to the overall operating strategy for the HOT lane, in particular how it works incombination with eligibility requirements, facility design and supporting technology. This factor considers

Lane management for priority or target users, andSetting the toll rate and eligibility requirements.

Can reasonable incident management be provided to assure travel time reliability?

Is there adequate maintenance support to assure quality service and operations, including all intelligenttransportation systems technology, flexible barriers, operation policy, and changes to service?

6

6

5

5

5

5

5

3

2

performance, and institutional issues, respectively. These tables illus-trate the depth and breadth of the number of factors that should beevaluated when considering the adaptation of an HOV lane to aHOT lane at the sketch-planning level. The default weights providedin each table indicate the relative importance of each of the factors.Factor weighting is described in more detail in the next section.

Factor Weighting

After narrowing the list of critical factors within each category tothose listed in Tables 1 through 3, researchers developed a weightingstrategy for the factors. The selected procedure ensures that factorsfrom all three primary categories (facility, performance, institutional)

Eisele, Burris, Wilner, and Bolin 71

are assigned weights and compared relative to each other categor-ically and globally. This also allows the analyst to assign weightsdifferent from the default values.

The High-Occupancy Toll Strategic Analysis Rating Tool (HOTSTART) program leads the analyst through screens related to eachcategory, allowing the analyst to input a weight for each factor, asshown in Figure 2. After addressing each factor in its respective cat-egory, the user is forced to make relative judgments of each factor,while maintaining a total summed weight of 100 for all factors acrossthe three categories (facility, performance, institutional). This waythe analyst has 100 “points” to allocate in a manner that he or shebelieves is appropriate for the community. This weighting strategyforces the user to consider simultaneously all the factors that areanticipated to affect the decision of whether to adapt from an HOV

TABLE 2 Performance Factors When Considering HOV Lane to HOT Lane Adaptation

Factor Description and Questions Addressed Default Weight

HOV lane utilization

Travel time savings/reliability

Public agency/societal benefits

Willingness to pay tolls

Safety

Environment

This factor examines actual usage (or predicted usage in the case of a planned HOV lane) of the HOV lane bynon-toll-paying vehicles from three viewpoints: Can the adaptation to a HOT lane remedy an existingutilization problem? This is measured by the number of non-toll-paying users of the HOV lane, and if a lackof such travelers causes the lane to appear “empty.”

Is there a potential that the increased use of the HOV lane will have a positive impact on GPL level-of-service?Will the adaptation have a positive impact on person-movement in the corridor? As long as HOV laneperformance does not deteriorate due to the additional toll paying vehicles, then at least some travelers(those toll paying vehicles) have improved trips while no travelers’ trips were worsened, thereforepositively impacting person movement in the corridor.

This factor examines both the amount of travel time savings offered by the HOT lane and the reliability of traveltimes on both the HOT lane and the GPLs. Like the lane utilization factor, the travel time factor will beexamined from three viewpoints:

Does the HOT lane offer significant travel time savings over the GPLs? This must include any additionaltime required for travelers to access the HOT lanes in the case where access is restrictive (as with the KatyHOV lane in Houston or I-15 express lanes in San Diego). This is a key consideration for adaptation as fewdrivers will pay for small travel time savings.Does adapting the lane to a HOT lane negatively impact the travel time on the HOT lane? If there is anegative impact—is it large and does it reduce the operating speed of the HOT lane below an agencyprescribed minimum acceptable speed?Are travel times on the HOT lane significantly more reliable (have less variability) than travel times onthe GPLs? Travelers will pay for additional reliability in their travel times. This measure must considerthe impact of incidents (crashes, stalls, etc.) on travel times for both the HOT lane and the GPLs. Theintent of measuring reliability is to ensure that a reliable trip is provided on the HOT lane at least 95% ofthe time. Example reliability measures include the 95th percentile travel time and Buffer Index.

This factor includes benefits of the HOT lane adaptation from both an agency revenue point of view and a netbenefit to society point of view. From the agency point of view, the greater the surplus toll revenue (totalrevenue minus start-up, operating, and maintenance costs) the better. From society’s point of view, anyoverall travel time savings, reduction in emissions, or reduction in fuel use are all benefits.

This factor examines local drivers’ willingness to pay tolls, both from their familiarity with tolls and theirincome levels. An interaction of these two issues will yield the appropriate scale values. Considerationsinclude

Are there other toll facilities already in the area? Do these other local facilities use the same tolltechnology as on the HOT lanes and will the transponders be interoperable?Travelers with higher incomes generally have higher value of travel time savings and are therefore morewilling to pay a toll to avoid congestion and reduce their total travel time.

This factor examines the likelihood that the adaptation will adversely affect safety on the HOV lane. Areduction in safety causes concerns for additional injuries due to the adaptation. Additionally, if there arefrequent crashes on the HOT lane then travelers will not pay to use the lane due to a fear of their ownsafety and the travel delays caused by crashes. The issue of safety is again relative to the city and corridor inquestion. However, for the scoring in this category, a significant decrease is a change in crash rate that issignificantly lower than the previous rate at a level of confidence of 95%. A slight reduction is a lower rate,but it is not statistically significant.

This factor includes impact of the HOT lane adaptation on both emissions and fuel use. Due to the highlikelihood that the adaptation will have minimal impact on either fuel or emissions, the default weight ofthis factor is relatively low. The minimal impact is caused by travelers in the (presumably congested) GPLsreducing some fuel use and emissions output by changing to the faster moving HOT lane, but travelers inHOV modes switching to HOT lane use as SOVs will increase the amount of fuel use and emissions output.

6

6

5

4

4

2

lane to a HOT lane. The analyst can change the weights by select-ing “adjust weights” and has the opportunity to save the new weight-ing profile. However, the sum of the weights must equal 100 beforethe user can continue.

Scoring Decision Trees for Factors

After addressing the factors and associated weights, researchersdeveloped a scoring method for each of the factors. Typically, the

72 Transportation Research Record 1960

scoring of any factor ranges from a value of 5 (highest) to −5 (low-est) with a score of zero generally indicating a minimally acceptablelevel for that factor. Decision trees were created for each factor toassist in scoring. The decision trees guide the analyst through perti-nent questions and issues for a given factor to determine the score.The score for each factor is then entered into the software tool. Thesoftware tool itself can guide a new user with questions that ultimatelyresult in the proper scoring.

Figure 3 illustrates a typical scoring decision tree for the facilitycross-section factor. It provides an example of how the user is guided

TABLE 3 Institutional Factors When Considering HOV Lane to HOT Lane Adaptation

DefaultFactor Description Weight

Public acceptance

Political acceptance

Environmental justice/Title VI issues

Revenue use

Interagency cooperation

Media relations

Sustained public education/information

This factor is concerned with public acceptability of adapting an HOV lane to a HOT lane or implementing a newHOT lane. The level of acceptability can be ascertained through focus groups or surveys. Additionally, public per-ception research can identify issues that are of importance to the public so that they can be addressed proactively.

This factor is concerned with the political knowledge of, and acceptability for, implementing a HOT lane. The politicalacceptance should be measured at all levels (e.g., local, regional, and state). Acceptance can be determined throughstakeholder interviews, supporting legislation, project champions, and media reports. Acceptance of HOT lanes canbe demonstrated by the adoption of such strategies into the long-range plan of an area and by enacting legislationthat allows for such adaptation projects. An adaption of an HOV lane to a HOT lane may also facilitate otherregional goals such as increasing person movement or increasing auto occupancy.

This factor concerns the disproportionate impact on low-income or minority populations that would be affected by aHOT lane. This may be different depending on whether the project proposes to adapt an HOV lane or to implementa HOT lane where none currently exists. This factor can be measured by the participation of affected groups in theplanning process and through focus groups or surveys.

There should be agreement prior to project implementation on the use of revenues derived from the project, if any.There may also be federal requirements that stipulate excess revenue use.

Interagency cooperation will be paramount to the success of a HOT lane. All agencies will need to support a adapta-tion project. Will multiple entities be responsible for maintenance and operation of the HOT lane? Operating agree-ments that are drafted may be required to stipulate certain provisions such as level-of-service or bus speeds perfederal regulations.

This factor deals with the media’s portrayal of the project. It may be influenced by an existing project or familiaritywith the HOT lane concept. It can be measured through editorials, media stories, and news clippings.

This factor concerns the mechanisms in place to generate support for a HOT lane project and the willingness to con-tinue public outreach after the project is implemented. Project success depends on the promotion of benefits theproject provides. Cross-jurisdictional support for project implementation is important to project success. Addition-ally, continued funding for advertising and outreach is needed.

6

6

6

5

4

2

2

FIGURE 2 HOT START screen shot of default weights.

to a particular score for the factor depending on the characteristicsof the corridor.

Factor Interactions

Once the final list of the most critical performance, facility, and insti-tutional factors were identified, it was necessary to investigate anypossible interactions between these factors. This would also affectscoring. For example, a poor (narrow) facility cross-section wouldhave a negative impact not only on the cross section factor but also onseveral performance factors. The narrow cross section could reducethe vehicle capacity of the lane, thereby reducing HOT lane utiliza-tion. It also could increase the crash rate, decrease average travelspeeds, and decrease a traveler’s willingness to pay to use the lane.After the factors from each area were examined, it was determinedthat those with the most impact were between the facility characteris-

Eisele, Burris, Wilner, and Bolin 73

tics and performance measures. Although both certainly can havesome interaction with institutional factors, those interactions wouldbe much smaller in magnitude and would make the analysis unnec-essarily complex without significantly affecting the outcome. There-fore, the remainder of this section, and the software itself, focus onthe interactions between performance and facility factors.

An argument can be made that almost any of the important facil-ity features listed in Tables 1 and 4 can, in some way, affect anyof the performance measures in Tables 2 and 4. It was the goal ofthis research and the accompanying software program to focus oninteractions that will have a material impact on the decision ofwhether or not to adapt an HOV lane to a HOT lane. To identifythese interactions, researchers first identified facility and perfor-mance measures with interactions that were ranked as follows: 1 = strong, 2 = moderate, and 3 = weak but still significant (seeTable 4). Second, researchers examined each of these interactionsas shown in Table 5. Finally, researchers adjusted the software

DesignenvelopesatisfiesAASHTOminimum

requirementsfor entirelength?

Length ofsections withunsatisfied

requirements:

No

Yes

< 100 ft *

5

Score

100 ft – 1000 ft *

1000 ft – 1 mile

1 mile to ½ facility

> ½ facility

Entire facility **

3

1

0

-1

-3

-5* Sections must be at least 1 mile apart.** This is a critical issue and upon scoring is noted in the Results.

FIGURE 3 Sample scoring decision tree for the facility cross-section factor.

TABLE 4 Interaction of Factors Affecting Adaptation of HOV Lane to HOT Lane

Performance Factor

HOV Lane Travel WillingnessFacility Factor Utilization Time to Pay Tolls Safety Environment Benefits

Cross section 1 3 2 1 When any of the

Lane separation 2 1 first five

Facility access for HOT O-D 3 2 performance factors

Facility access design 2 2 1 are impacted,

Ability to enforce 3 3 the benefits

Facility signage 2 2 3 of the HOT

Pricing strategy 1 1 lane are impacted.

Incident management 3 3 2 3 3

Maintenance 3 3 3 3

1 Strong interaction2 Moderate interaction3 Weak, but significant interaction

Secondary interaction (“benefits” column only)

package so that these interactions were accounted for in the finalHOV to HOT rating.

These strong, moderate, and weak interactions are accounted forin the software by first obtaining the relevant facility characteristic(e.g., cross section) score from the user. If the value of the charac-teristic is less than ideal, then some adjustment of the related per-formance factor (e.g., lane utilization) is required because thedefault performance factor values assume an ideal facility. Thesoftware automatically updates the performance factor to reflect thissuboptimal facility characteristic by subtracting a set number ofpoints from the value of the performance factor. While the numberof points subtracted varies by interaction type and strength of theinteraction, typical reductions are one to two points.

After a final factor score and its weight are determined for eachfactor, the two values are multiplied together to get a total value forthat factor. The sum of these factor values in each category (facility,performance, institutional) provides a category score. The sum of thethree category scores provides the overall score for the analysis. Thescoring is illustrated in an application at the end of the paper.

HOT START SOFTWARE

The HOT START software is a Microsoft Windows–based programthat is built on the Visual Basic.NET platform (Microsoft Corpo-ration, Redmond, Washington). The software guides the analyst

74 Transportation Research Record 1960

through the process of evaluating an HOV facility for possibleadaptation to a HOT facility, following the guidelines discussedin this paper.

The software provides the full functionality of a MicrosoftWindows–based program, including the ability to save, load, print,copy, and provide access to various help functions. The softwarealso ensures a mathematically accurate analysis by automating theinteractions between various factors as well as leading the analystthrough a series of steps and questions to obtain the correct score fora given factor, as shown in Figure 4. For various factors, additionallinks are provided to documents, websites, and telephone numbersthat will further help the analyst answer questions to determine theappropriate score. In Figure 4, the helpful links include maps show-ing the distribution of income across specific cities, attached .pdffiles with step-by-step instructions on how to find the informationneeded, and an Internet link to the U.S. Census Bureau website.Additionally, a data collection form is available in HOT START’sHelp menu for the analyst to use to gather the necessary informationto be input into the software.

When starting the software, the analyst is first presented with atitle screen, which enables an existing analysis file to be loaded or anew analysis to be started. If a new analysis is selected, the analystis asked whether default weights should be used. If default weightsare not selected, a dialog screen appears that allows the analyst todistribute 100 points (weights) among the factors as describedpreviously and shown in Figure 2.

TABLE 5 Discussion of Significant Interactions Between Factors

Facility Factor Performance Factor Interaction Discussion

Cross section

Lane separation

Facility access for HOTlane origins and destinations.

Facility access design

Ability to enforce

Facility signage

Pricing strategy

Incident management

Maintenance

HOV lane utilization

Travel time

Willingness to pay tolls

Safety

Travel time

Safety

HOV lane utilization

Willingness to pay tolls

Travel timeWillingness to pay tolls

Safety

Willingness to pay tolls

Willingness to pay tolls

HOV lane utilizationWillingness to pay tolls

All

All except environment

As the cross section narrows, the volume of vehicles accommodated on the lane at free flow speedsdecreases.

As the cross section narrows, the free flow speed drops, decreasing the travel time benefits of the HOVlane.

With very narrow lanes travelers may not feel comfortable and safe in the lanes, decreasing theirwillingness to pay for travel in those lanes.

Both actual and perceived safety may decrease as lane widths decrease. Increased crashes will alsoadversely impact travel times. Additionally, if insufficient room exists to move stalled or crashedvehicles out of the travel way on a barrier-separated lane then travel times could be much worse thanon the GPLs.

If a significant blockage occurs in a barrier-separated facility (frequently) then travel times on the HOVlane will deteriorate significantly.

Limited research suggested barrier-separated lanes to be safer than lanes separated by a buffer or aflexible barrier.

If the access points for toll paying drivers are congested, the number of non-paying travelers atthose access points will decrease.

If the access points for toll paying drivers are congested or located long distances from their preferredentry point, the travel time savings offered by the HOV lane is reduced.

Poor access/egress points can add travel time to the HOV lane option.Poor access/egress points can impact travel time to/from the HOV lane, reduce perceived/actual

safety, and reduce ease of use, which all impact the driver’s willingness to pay for the lane.Poor access/egress points can reduce perceived/actual safety.

Some potential paying customers may choose to be violators if they perceive/recognize lax enforcement.

Adequate pricing/occupancy requirement information must be available before many travelers electto pay for HOV lane use.

The pricing strategy clearly has a major impact on both the utilization of the lane and the traveler’swillingness to pay the toll. The software provides guidance on the preferred pricing strategy for dif-ferent lane options and assumes the HOV lane operator selects an appropriate strategy.

An aggressive incident management strategy that rapidly clears incidents from the HOV lane canimprove all performance aspects.

If there is debris in the lane on a regular basis, or there are issues with reversing a reversible lane, thenthis will impact several aspects of HOT lane performance.

Once the preliminary weighting of factors has been accom-plished, the analyst is presented with a series of screens. If the ana-lyst loaded an existing analysis file, then the saved informationappears on the screen. Otherwise, for new analyses, the text boxesare blank. The data required on the first screen include general infor-mation about the analyst and the facility being evaluated. After gen-eral information is entered, the analyst proceeds to the facility,performance, and institutional screens to enter and adjust weights ofthe various factors. After weights are entered, the user proceeds toscore the factors.

The score for any factor ranges from −5 to +5. Help in deter-mining the score for the facility under investigation can be obtainedby clicking the hyperlink text associated with that factor. The ana-lyst will then be guided through a series of steps and questions asshown in Figure 4 to obtain the score, or the final factor score canbe entered directly into the score box as identified with the circled“1” in Figure 5.

The software provides several other useful functions on thesescreens to help with the analysis process (see Figure 5). First, whenthe user rests the mouse over a factor, a brief factor descriptionappears in a textbox. This is identified in Figure 5 with a circled “2.”When the mouse pointer is moved over the score box, a descriptionof that score based on the particular factor is shown, to ensure theanalyst entered the desired score on the basis of factor conditions.

Eisele, Burris, Wilner, and Bolin 75

This feature is identified with the circled “3.” As identified with a cir-cled “4” in Figure 5, the analyst can tag a specific factor as “unsure,”which will be noted in the results. Critical factors that occur on thebasis of certain factor scores are flagged automatically by the soft-ware and are noted in the results. Finally, a colored meter at the bot-tom of each factor category gives the analyst an idea of the expectedresults for that particular category. This is identified in Figure 5 witha circled “5.”

In this example, the performance category score is below zero,indicating relatively poor results. While gray scale is shown in the fig-ures in this paper, the facility, performance, and institutional metersare color coded in the program. At the negative end of the meter spec-trum, red indicates the scores are poor; at the positive end of the spec-trum, green indicates the scores are relatively positive. When shownin color, the performance meter in Figure 5 is yellow.

After the analyst has finished entering the scores and weights, asummary page, which shows the factors in order of weight value,can be viewed. At this point, weights can be adjusted, if necessary.If the analyst is satisfied, the results of the analysis can be computedand viewed. The results pages can be viewed and printed. The resultsprovide two key components of information: (a) scores and (b) crit-ical factors to be resolved. Information on the results pages includesthe overall score (resulting scores page), individual category (facil-ity, performance, institutional) assessments (resulting scores page),

FIGURE 4 Prompts for selecting appropriate factor score.

critical factors (remaining critical factors page), and factors thatwere marked as “unsure” (remaining uncertainties page). The inter-pretation of HOT START’s results is further illustrated through theapplication in the next section.

HOT START APPLICATION: I-10 EXISTING HOV LANE, HOUSTON

The I-10 (Katy Freeway) HOV lane has been open since 1984. Con-tinuing increases in the number of HOV lane travelers caused thelane to become congested during peak hours, so the peak hour occu-pancy requirement was raised to HOV 3+ in 1988. To improve over-all efficiency, TxDOT and the Metropolitan Transit Authority ofHarris County (joint operations partners) considered the HOT laneoption during 1997 and later implemented the HOT lane in 1998.This case study is based on the conditions in place in 1997, when theinitial evaluation would have occurred.

The default weights used and the assigned scores for each factorare shown in Figure 6. The facility meter at the bottom of Figure 6indicates that the overall score for the facility factor is acceptable.Though shown in gray scale in Figure 6, the scale is green whenshown in color—representing relatively positive facility results. The

76 Transportation Research Record 1960

only low score (−2) is for the ability to enforce. All the other factorshave positive scores.

The default weights and assigned scores for the performance fac-tors are shown in Figure 7. The small number (−1) between the scoreand weight columns represents the amount of points deducted fromthe score for that factor (willingness to pay tolls) to account for inter-actions as discussed previously. The performance meter implies thatthe score is positive (green if shown in color).

Like the bars for facility and performance considerations, thebar for institutional considerations indicates positive-scoring fac-tors (again, it would be green if shown in color). The results for theinstitutional factors are displayed in Figure 8. All factors scored at or above zero except for public education/information, which scoreda −3. However, because of the low weight of this factor and thehigher scores of the other factors within institutional considerations,the −3 score hardly affected this category.

The graphical results of the analysis for I-10 are shown in Figure 9(resulting scores page—see lower left of Figure 9). In this example,there are no critical factors to be addressed. In the program, had therebeen critical factors, a red circled “x” would appear adjacent to theappropriate category column. The remaining critical factors page (tabat lower left of screen) would then summarize the critical factorsidentified in the analysis. In this analysis, none of the factors was

1

24

5

3

FIGURE 5 Screen for entering scores and weights into HOT START.

Eisele, Burris, Wilner, and Bolin 77

FIGURE 6 I-10 HOV facility factors weights and scores.

FIGURE 7 I-10 HOV performance factors weights and scores.

78 Transportation Research Record 1960

FIGURE 8 I-10 HOV institutional factors weights and scores.

FIGURE 9 I-10 HOV resulting scores page.

marked as “unknown.” Therefore, no such items are identified on theremaining uncertainties page (tab at lower left of screen) of theresults. As shown in Figure 9, the maximum possible score, based onthe default weighting profile used, is 210 for facility, 135 for perfor-mance, and 155 for institutional. The actual scores for the individualcategories are 140, 48, and 75, respectively. Therefore, quantita-tively, the potential adaptation to a HOT lane results in relativelypositive results for the facility, performance, and institutional cate-gories. The overall score for the project is 263. It should be notedthat the overall value can be used to compare projects in which thesame weighting scheme has been used.

The I-10 analysis confirmed the decisions made several years ear-lier to proceed with adapting the HOV lane to a HOT lane. Cur-rently, the facility is being reconstructed to include two managedlanes in each direction.

CONCLUSIONS

This paper documents research that provides an analytical frame-work (HOT START program) to assess the critical factors thatshould be examined when considering the adaptation of an HOVlane to a HOT lane. It allows analysts to determine quickly thoseHOV lanes that, when adapted to HOT operations, have a high prob-ability of successfully meeting several key goals. This then allowsagencies to focus detailed analyses (such as a benefit–cost calcula-tion) on facilities that most deserve the additional analytical effort.Key conclusions of the research are provided below.

Compilation of Key Factors

The research provides the first attempt in the literature of an ana-lytical tool for assessing critical factors before the adaptation of anHOV lane to a HOT lane. In particular, the research provides a frame-work for the consideration of factors that relate to key facility, per-formance, and institutional factors in a diagnostic software tool thatcan be tailored to the specific needs of the community in which theparticular project is located.

Decision Tree Scoring

The HOT START tool provides a unique method of scoring eachfactor with decision trees. The decision trees guide the analyst to theappropriate score (−5 to +5) by answering questions related to eachfactor. The decision trees are based on the latest research on HOV,HOT, and managed lanes. The decision trees also provide a methodof scoring relatively qualitative factors.

Interaction Effects

The interactions between factors are also considered in the softwaretool. For example, facility design affects facility performance, and

Eisele, Burris, Wilner, and Bolin 79

HOT START provides an analytical way to consider and includethese interaction effects.

Case Study Application

The HOT START analytical tool is applied to the case study ofI-10 (Katy Freeway) in Houston, Texas. The example illustrateshow the tool can be applied to evaluate the facility, performance,and institutional factors of interest when considering adapting anHOV lane to a HOT lane. The sample case study illustrates how anoverall score and critical factors can be identified with a real-worldexample.

Flexibility

As more research becomes available related to any of the key fac-tors considered in HOT START, the factors (scoring, interactions)can be updated in future versions of the software.

ACKNOWLEDGMENTS

This paper is based on research sponsored by the Texas Depart-ment of Transportation. It was performed by the Texas Trans-portation Institute of the Texas A&M University System. Theauthors would like to thank the Texas Department of Transporta-tion for sponsorship of the research on which this paper is based.The authors would also like to thank William Stockton, GingerGoodin, Tina Collier, and Michelle Hoelscher of the Texas Trans-portation Institute, who provided technical insight to the TxDOTresearch effort on which this paper is based and provided theircomments on this paper.

REFERENCES

1. Projects. Value Pricing Homepage. www.valuepricing.org. AccessedJuly 29, 2005.

2. Guide for High-Occupancy Vehicle (HOV) Facilities. AASHTO,Washington, D.C., 2004.

3. Fitzpatrick, K., M. A. Brewer, and S. Venglar. Managed Lane Ramp andRoadway Design Issues. Research Report 0-4160-10. Sponsored by TexasDepartment of Transportation. Texas Transportation Institute, CollegeStation, Jan. 2003.

4. Chrysler, S. T., A. Williams, S. D. Schrock, and G. Ullman. Traffic ControlDevices for Managed Lanes. Research Report 0-4160-16. Sponsoredby Texas Department of Transportation. Texas Transportation Institute,College Station, April 2004.

The contents of this paper reflect the views of the authors, who are responsible forthe facts and the accuracy of the data presented here. The contents do not nec-essarily reflect the official views or polices of the FHWA or the Texas Departmentof Transportation.

The Congestion Pricing Committee sponsored publication of this paper.