Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
10 This chapter discusses the categories of benefits and disbenefits associated with dedicat- ing lanes to CAVs. The discussion includes: (1) types of stakeholders who will be impacted directly by dedicating lanes to CAV users, (2) factors influencing the benefits and disbenefits, and (3) categories and lists of performance measures that represent these benefits and dis- benefits. Figure 2.1 provides a summary of this discussion. 2.1 Types of Stakeholders Dedicating lanes to CAVs directly impacts three categories of stakeholders: ⢠DL users. Depending on which vehicles are allowed on the DLs, these stakeholders could include CVs, AVs, or even HOVs and tolled single-occupancy vehicles (SOVs) in the case of high occupancy toll (HOT) lanes. ⢠GPL Users. These stakeholders include all CAVs and HOVs that do not use the DLs, as well as the unequipped SOVs that are not allowed on the DLs. ⢠Facility Owners and Operators. These stakeholders will see a shift in lane/facility usage as well as other performance measures. 2.2 Factors Influencing Benefits and Disbenefits Based on a preliminary analysis, the impacts on each of these categories of stakeholders depend on factors pertaining to the DLs as well as the CAV technology and market penetration. The following subsections describe these influencing factors. 2.2.1 CAV Market Penetration The CAV MPR represents the percentage of CAV users in a traffic stream at a given location. When evaluating various lane-use strategies, it is critical to estimate the potential economies of scale that attend CAV MPRs. Basically, the more people choose to use CAVs, the better the overall gain in corridor efficiency. Ongoing research has shown that incorporating DLs can make a significant difference in cases of low CAV MPR (e.g., less than 30%) (Talebpour et al. 2017). Moreover, the benefits to three user typesâCAV users, GPL users, and facility owners/ operatorsâincrease as the CAV MPR increases. As more vehicles move to the DL, the level of service (LOS) of the GPL increases. Potentially, the GPL can achieve a drastically larger carry- ing capacity, resulting in improved overall system performance from the perspective of facility operators/owners. The CAV MPR can be a significant factor in determining the total benefits/ disbenefits for different users. C H A P T E R 2 Categories of Benefits and Disbenefits to Stakeholders
Categories of Benefits and Disbenefits to Stakeholders 11 2.2.2 Roadway Geometry Roadway geometric configurations are likely to affect the impacts of CAV DLs. These configu- rations may include physical barriers, access points, length of the facility, multi-lane treatments, and shoulder use. The following sections briefly explain each of these configurations. 2.2.2.1 Physical Barriers In some instances, physical separation of a managed lane may be necessary to ensure higher performance for the DL. Physical barriers may be needed when: (1) the prevailing speeds of the DL and GPL are significantly different; (2) undesirable weaving movements between the DL and GPL need to be limited; and (3) instances of toll evasion need to be prevented. The phenomenon by which the speed of traffic on the DL is affected by the speed of traffic on the GPL is termed the frictional effect; the degree of congestion in the GPL impacts the speed of the managed lane (Fitzpatrick et al. 2016b). This study also found that the placement of soft but visible barriers (e.g., pylons) mitigated the influence of the GPL speed on managed lane operation: for an increase in speed of 1 mph on the GPL, the increased speed on the man- aged lane was 0.42 mph without pylons and 0.3 mph with pylons (Fitzpatrick et al. 2016b). For DL users, the primary benefit of a physical barrier is minimized impact of the GPL speed on the DL. However, a physical barrier also can have disbenefits. For example, the physical barrier reinforces the âsnailâ effect, whereby the slowest moving vehicle in an HOV lane can govern the speed of the entire lane (Kwon and Varaiya 2008). Furthermore, a physical barrier would produce disbenefits such as requiring an additional construction budget for building the barrier and potential delays for accident clearance due to limited access to the lane. Dedicating Lane(s) for CAV Users Stakeholders Dedicated Lane Users General Purpose Lane Users Owners and Operators of the Facility Factors Influencing Benefits and Disbenefits CAV Market Penetration Roadway Geometry Enforcement Intensity Operation Hours CAV Technology Functional Types: (a) CAV Only, (b) CAV+HOV, (c) CAV+HOV+HOT Toll Collection Attributes Performance Measures Mobility Energy and Environmental Safety Societal Figure 2.1. Benefit/disbenefit categories, influencing factors and affected stakeholders.
12 Dedicating Lanes for Priority or Exclusive Use by Connected and Automated Vehicles 2.2.2.2 Access Points In conjunction with the common trip characteristics, the number and locations of DL access points need to be considered thoroughly. Classification of a DL as having unlimited access or restricted access affects the laneâs design and operational characteristics. An inappropriate or poorly planned design for access points would likely waste investment funds, downgrade the performance of the DL, and erode public support for the DL. Furthermore, the number and locations of the access points could affect enforcement and even the safety of a DL. Continuous access is a common approach (and the least expensive approach) for part-time operation of a DL. Restricted access using features such as access zones or ramps generally is more expensive to implement and typically is reserved for large concurrent-flow or reversible facilities. With restricted access DLs, the major benefits for both DL users and GPL users are (1) a reduction in the number or length of road sections that involve possible weaving and (2) the same positive effects as those mentioned under physical barriers. The disadvantage of a restricted access DL is reduced safety at locations approaching the access points. Weaving motions can become concentrated at such locations as drivers experience more pressure to complete lane changes than would be the case with continuous access. Operating agencies need to evaluate closely the return on investment for constructing restricted access points in comparison to the performance trade-off for using a continuous access point. 2.2.2.3 Length of the Facility It is suggested that the length of the managed lane be carefully considered and constructed based on the regional traffic patterns. The length of the DL could spark potential equity issues. For example, given the high investment cost of DLs, the owner can be expected to be sensitive to the return on investment. Greater benefits can be expected for CAVs with a longer facility. 2.2.2.4 Multi-Lane Treatments When warranted by the intended demand and given available space, developing a DL policy that incorporates multiple lanes is an effective way to scale up the DL deployment. Kwon and Varaiya (2008) showed that the capacity of a single-lane DL could be reduced by 20% due to the so-called âsnailâ effect, when a faster vehicle cannot pass the slower one in front of it. Eliminating the snail effect and increasing capacity requires the DL to have more than one lane. The 14.5-mile, 2-lane HOV segment on SR-91 in Orange County, California, does not seem to suffer from the snail effect. The Port Authority of New York and New Jersey (PANYNJ) is considering adding a second bus-exclusive lane on I-495 with the option of passenger-vehicle usage on payment of an additional toll (Hess et al. 2011). 2.2.2.5 Shoulder Use Shoulder use is an important influencer for DL management. For example, a Bus on Shoulders (BOS) policy was implemented in 2006 by the New Jersey Department of Trans- portation (New Jersey DOT) as part of the Enhanced Bus Improvement Program. The origi- nal shoulder of a 6-lane arterial highway was improved by adding full-depth pavement for buses on the shoulder. A speed limit of 35 mph was imposed on the shoulder, which is reserved for buses from 5 am to 9 am during weekdays. Approximately 440 buses and 6,800 passengers were served daily during the peak period when BOS is in effect (Martin et al. 2012).
Categories of Benefits and Disbenefits to Stakeholders 13 2.2.3 Enforcement Intensity Enforcement of the proper use of DLs is vitally important, regardless of whether the vehicles are operated using manual or automated methods. For CAV users, enforcement ensures both a high LOS on the DL and greater safety, as it reduces the likelihood that GPL vehicles will cut into the DL. Regardless of the enforcement method chosen, however, enforcement comes with a cost. Consistent, effective enforcement confers the greatest benefits when significant differences exist in the LOS between the GPL and the DL. In the context of CAV DLs, a system that can discern GPL vehicles from CAVs needs to be developed and implemented. Additionally, the facility design should be conducted with enforcement in mind. For CAV DLs, the primary enforce- ment task involves confirming DL user eligibility, by detecting the CAV equipment of individual vehicles (e.g., the DSRC onboard unit). This enforcement task can be performed easily through roadside equipment deployed along the lane. If the DL allows HOV/HOT users, however, it can be challenging for roadside equipment to detect in-vehicle passengers. To enhance DL enforce- ment and reduce violations, suggested actions include: ⢠Conducting engineering meetings with state highway patrol officials to determine roadside enforcement locations; ⢠Introducing prominent roadside signs that specify the consequences of violations; and ⢠Adopting methods for conducting random visual and special enforcement that can be deployed in addition to routine enforcement methods (FHWA 2016a, FHWA 2016b). 2.2.4 Toll Collection Technology and Toll Pricing Methods Electronic toll collection technology makes variable pricing simple and easily accomplished. The deployment of DSRC and other technologies (e.g., license plate capture) can further auto- mate enforcement tasks. Several products are market-ready, but more research is needed with regard to developing technologies that can automate enforcement for all possible operating scenarios, including cost-effective, large-scale deployment of automated passenger occupancy detection technologies. Enabling legislation also may be warranted (McCune 2015). Toll collection information (e.g., pricing, travel times) also could impact adoption and usage of DLs. Ideally, CAV users could receive the toll collection information directly into their cars, potentially reducing the need for expensive overhead signage. 2.2.5 Operation Hours The hours of operation for managed lanes typically are determined by traffic congestion con- ditions (Fisher 1995, Dahlgren 2002, Kwon and Varaiya 2008, and Avelar et al. 2016). During highly congested periods (such as a.m. and p.m. peaks), access to the managed lanes is restricted to the users designated by the owner/operatorâs policies; during other time periods, the lanes typically are opened to all GPL users. The hours of operation for managed lanes can affect the performance of one or several travel lanes dramatically. Careful consideration of the hours of operation of managed lanes can minimize potential adverse effects of the managed lanes on the rest of the traffic. 2.2.6 CAV Technology Types and Implementation CAV technology and applications are evolving rapidly. The U.S.DOT has conducted signifi- cant research on a wide range of CV applications. These applications include both V2V and V2I technologies. The specific CAV technology implemented will have a huge impact on the benefits and disbenefits achieved.
14 Dedicating Lanes for Priority or Exclusive Use by Connected and Automated Vehicles 2.2.7 Functional Types Given low MPRs during the initial stages of dedicating lanes for CAVs, CAVs may need to share the DL with other users, or even GPL users (e.g., in the case of HOV/HOT lanes). Various functional DLs can be listed based on their user types, as follows: ⢠CAV DL. A lane that can be used by both CV and AV groups; ⢠CAV + HOV Lane. A lane that allows CVs, AVs, and HOVs to travel, making this type of lane one of the viable transition options to promote both CV and AV usage with existing HOV lanes; and ⢠CAV + HOV and HOT Mixed-Use Lane. A lane that gives priority for CAVs, HOVs, and HOT vehicles with the goal of moving as many passengers as possible on the managed lane. The demand for HOV travel heavily affects the success of HOV lanes, as demonstrated by multiple case studies shown in Murray et al. (2000). The same logic can be applied to other DL strategies. Underutilization of a DL not only creates a negative impression of DLs among GPL users, it also can worsen the mobility performance of the GPL if more vehicles are forced to use the GPL because of ineligibility to access the DL. Since the passage of the Transporta- tion Equity Act for the 21st Century (TEA-21) in 1998, broadening access to HOV/HOT lanes for low-emission and energy-efficient vehicles when available capacity exists has expanded the user group for these DLs. As of October 2013, 13 states had adopted provisions that allow special groups of vehicles to use HOV/HOT lanes. Some states issue special license plates, stickers, permits, or decals to identify the exempted vehicles, and some states charge a small fee (e.g., $8 per use in California) (Turnbull 2014). For example, a physically separated managed lane facility was initiated on the reversible 2-lane stretch of I-15 in northern San Diego County, California. The managed lane was designed initially for use only by HOV traffic, but the policy was adjusted to allow SOV users to pay to travel on the HOV lane. The fee for using the HOV lane varies depending on the level of congestion of the corridor. Information to motorists is provided via changeable message signsâalso called dynamic message signs or vari- able message signs (VMS)âthat can be updated as frequently as every 3 minutes (Brownstone et al. 2003). Taking a similar approach to CAV DL implementation, CAV users could preserve a higher right-of-way on the DL. Such a policy would attract more CAV demand into the DL. As a result, GPL users could benefit from reduced CAV demand on the GPLs, which improves the LOS for the GPLs. In turn, the owner/operator could retain public support for the DL policy being implemented. 2.3 Performance Measures Increasingly, CAV applications have gained attention due to their potential to enhance traf- fic operations fundamentally. To implement the managed lane strategies for CAV technologies successfully, the research team grouped the relevant factors into four subcategories for dedicat- ing lane(s) to CAVs (see Figure 2.2). Details on each subcategory are discussed in the balance of this section. 2.3.1 Mobility Mobility performance measures can be used to assess improvements in the ease of moving travelers and vehicles in a network. These measures can include origin-destination patterns, daily peak hours, recurrent congestions, and so forth. When implementing managed lane strat- egies, it is important to ensure that the strategies are suitable for the traffic flow characteristics.
Categories of Benefits and Disbenefits to Stakeholders 15 Four major factors are associated with mobility: level of congestion, travel time reliability, aver- age speed, and vehicle occupancy. 2.3.1.1 Level of Congestion On most occasions, the level of congestion of a roadway or a corridor provides justification for the authority to implement appropriate managed lane strategies. Dahlgren (2002) studied the relationship between the effectiveness of a managed lane and congestion level of a roadway, concluding that adding a mixed-flow lane is more effective in delay reduction than an HOV or HOT lane when initial maximum delay is 30 minutes or less. If initial delay and HOV MPR are very high, however, an HOV lane would become more effective than either a mixed-flow or HOT lane (Dahlgren 2002). Dedicating lanes to CAV users may alleviate traffic congestion in a simi- lar way. DL users gain benefits from the increased mobility under heavy congestion, GPL users remain at the same LOS, and the operator likely gains the benefit of a more productive facility. 2.3.1.2 Travel Time Reliability In measuring both recurrent and non-recurrent congestion for a specific roadway, travel time reliability factors indicate the extra time a traveler has to allocate for likely on-time arrival. Value of time (VOT), along with value of reliability (VOR), are primary indicators for whether a road user will pay to use a managed lane (i.e., a HOT lane). As discussed in the I-394 MnPASS case, the VOT is the monetary value that a managed lane user places on the reduction of travel time, whereas the VOR is the monetary value that the user places on improvement of travel time predictability (Carrion and Levinson 2012, Janson and Levinson 2014). 2.3.1.3 Average Speed One operational goal of DLs is to prevent demand from reaching the roadwayâs carrying capacity, because the stability of the flow deteriorates as the demand of a roadway approaches its capacity. Unstable flow is likely to result in delays at best and in breakdown conditions at worst. A study in California by Kwon and Varaiya (2008) showed that HOV lanes typically achieve 1,600 vehicles per hour per lane (vphpl) at 45 mph speed, and that throughput is not linearly Figure 2.2. Performance measures for the benefits and disbenefits of various users. Mobility SafetySocietal Justice Energy and Environment Performance Measures Level of Congestion Travel Time Reliability Average Speed Vehicle Occupancy Crash Rate Speed Fluctuation Fuel Consumption Emission Perception of Exclusivity Public Outreach Equity
16 Dedicating Lanes for Priority or Exclusive Use by Connected and Automated Vehicles related when demand is close to carrying capacity. To harness the optimal cruising speed for the DL, optimal average speed needs to be determined after considering the trade-offs with mobility. CAVs could achieve smaller headways using applications such as CACC systems, which could enable free-flow speeds to be maintained even at higher traffic volumes. 2.3.1.4 Vehicle Occupancy Vehicle occupancy is less influential on performance outcome when the MPR for CAVs is low. When the MPR becomes high, however, depending on the CAV lane-use strat- egies, change in occupancy could alter the outcomes for performance. Thus, the vehicle occupancy rate can be a direct indicator to examine the impact of CAV technologies for both GPL and DL users, depending on the lane management policies adopted by the operator (e.g., HOV 2+, HOV 3+). 2.3.2 Safety Traffic safety is a crucial factor in the implementation of CAV technologies. In this study, the two performance measures considered for safety were crash rate and speed fluctuation/ difference. 2.3.2.1 Crash Rate The crash rate is a reliable representation of roadway safety. It is the ratio of the number of crashes at a given period to the vehicle-miles traveled (VMT). Combined with appropriate man- aged lane strategies, CACC has the potential to reduce the overall crash rate at an implemented roadway. Given that managed lane strategies can provide CACC-equipped vehicle groups with an exclusive right-of-way, a homogeneous traffic stream could be achieved in the managed lane, thereby resulting in traffic conditions that reduce the human driversâ interference with vehicle maneuvers. CAV applications also could provide enhanced information and faster reaction times to the CAVsâ âdriving systemsâ in order to reduce crash rates. For example, CAV appli- cations could provide advance information about congestion queues in order to reduce the frequency of secondary crashes that occur when fast-moving vehicles suddenly approach much slower or stopped vehicles. 2.3.2.2 Speed Fluctuation/Differences Given that crash statistics rely on a large amount of data, collecting data on direct safety per- formance measures such as crash rates would have been challenging for this study. To overcome this issue, surrogate safety measures were applied to examine the safety impacts of transporta- tion system improvements. Unfortunately, the surrogate measures that have been developed have only been validated for human-driven vehicles, not for vehicles that are partially (or com- pletely) driven by computer. As a result, and reflecting the differences in the factors that deter- mine traffic safety for CAV versus human-driven vehicles, these surrogate measures can only be applied to CAV systems with extreme caution. Among the various surrogate safety measures, speed fluctuation is one of the primary fac- tors for the occurrence of both initial and secondary traffic crashes. Speed fluctuation can be measured in terms of the difference in 95th percentile spot-speeds of vehicles, between lanes, segments, or even time frames. After implementing CV, AV, and CV and AV DL strategies, any changes in speed fluctuation should be studied to quantify the actual effects of these strate- gies on safety performance. The implementation of CV, AV, and CV and AV DL strategies will likely reduce speed fluctuations (and therefore improve traffic safety), but to date measurement of speed fluctuation has been conducted indirectly, using data collection technologies such as
Categories of Benefits and Disbenefits to Stakeholders 17 inductive loop detectors, radar sensors, and probe vehicles (Gettman et al. 2003). For imple- mented CAV DL lanes, the existing approaches to measure speed differences can be retained while CV technologies enable the capture of individual vehiclesâ speed profiles. Speed informa- tion could be obtained from basic safety messages from CVs equipped with DSRC. Moreover, while it is not mandatory to share the data, AVs that are capable of detecting adjacent vehicle movements could produce high-fidelity local speed fluctuation data. 2.3.3 Energy and Environment Frequent braking and acceleration has been shown to be energy inefficient and can increase emissions (Rakha and Ding 2003). Specifically, energy is lost through decreased momentum when vehicles brake and increased fuel is used to create the energy needed when accelerating to return to the original speed. CAVs on a DL can mitigate the impacts of braking and accel- eration because the CAVs can coordinate and smooth out their speed changes, accelerating and decelerating in an automated manner. Kall et al. (2009) used the MOBILE-Matrix model to predict the change in emissions resulting from the conversion of an HOV lane to a HOT lane on I-85 in Georgia. Four pollutantsâcarbon monoxide, hydrocarbons, oxides of nitrogen, and particulate matterâwere considered in the study, which found that the vehicle emissions are influenced by change in VMT, speed, and vehicle fleet characteristics (Kall et al. 2009). Similar model-based approaches could be applied to estimate emissions and fuel consump- tion. Furthermore, the CAV environment with sufficient MPRs will enable direct collection of emission and fuel consumption from CAVs to improve the quality of emission and fuel consumption models. 2.3.4 Societal Justice It is necessary to consider the societal impacts of CAV technologies. This section describes three performance measures covering the perspectives of social justice and public relationships. 2.3.4.1 Perception of Exclusivity The concept of dedicating some lanes for use by particular vehicle groups is philosophically debatable and could potentially create or aggravate social justice bias related to incomes and social class. Depending on the strategy and policies applied to create a CAV DL, some groups of people may not be able to afford access. Studies of the willingness to pay for HOT service con- cluded that high-income users were more likely to consider alternative travel means, compared to low-income users (Finkleman et al. 2011). 2.3.4.2 Equity Approaches to equity issues can be considered in developing DL strategies for combinations of CV, AV, HOV, and HOT users. Possible considerations include: ⢠Providing GPL vehicle users with subsidies or cost incentives to expedite the transition from GPL vehicle to CAV; ⢠Ensuring the presence and designation of redundant/alternate roadways so that users have choices beyond the GPLs; and ⢠Conducting DL design with consideration for the limited availability of and accessibility to CAV lanes for low-income people. Agencies can involve individuals with low income in the planning process by providing vari- ous incentives (e.g., partial tax relief, monetary subsidies, educational advantage) depending on
18 Dedicating Lanes for Priority or Exclusive Use by Connected and Automated Vehicles the situation of the community. Addressing potential equity and environmental justice issues often is vital to obtaining the support needed to implement and design DL projects. It is impor- tant to consider equity at all stages of the project to ensure a successful implementation and establish the foundation for future similar projects in the region. Equity issues should be ana- lyzed and presented proactively and as early as practical. It can be difficult to gain public support, especially if perceived equity issues are not adequately considered. Some general guidelines have been used by public agencies to address similar concepts (e.g., road pricing projects), and these can be considered when developing CAV DLs. Despite pervasive references to equity, consensus on the definition of equity has not been reached, because the possible dimensions for what constitutes an equitable policy depend on the context and community priorities. Regarding equity for DLs, key concerns include: ⢠Whether it is reasonable and acceptable for certain groups to disproportionately receive positive impacts (benefits) or negative impacts (disbenefits) from the DL, ⢠How large the differences are between the benefits received by favored groups and the disbenefits received by disfavored groups, and ⢠What compelling public policy objective is served by introducing the inequity (Weinstein and Sciara 2004). The most common concern regarding DLs is affordability for low-income groups to use the facilities or technologies. With HOT lanes, it was observed that low-income groups of users used the service depending on the importance of a trip (FHWA 2008a); however, users of CAV technologies are likely to have higher incomes in areas of low MPR. The necessary initial invest- ments for acquiring CAV technologies make it challenging to use the HOT lane case as a model for implementation of a CAV DL. Geographic equity is important with respect to the spatial patterns or apparent segregation in accessibility for different constituencies. In the case of Marylandâs US-50, it was argued that suburban commuters with higher incomes received disproportionately more benefits from the HOT lane, compared to residents who lived near downtown and had no access to the lanes (Weinstein and Sciara 2004). It is not uncommon that the geographic equity implications change as a project expands or changes. Modal equity should not be overlooked, especially when the operator would like to promote certain modes of transportation (e.g., HOV, transit, CAV) and provide incentives to users for choosing one mode of travel over another. Similarly, when considering the improved mobility of goods, which directly affects the nationâs economic growth, truckers should be considered in the distribution of CAV DL benefits from modal equity perspective. The Minnesota legisla- ture decided to direct 50% of HOT lane revenues in excess of project cost on I-394 to transit in the corridor. Some advocacy groups proposed that the pricing on the HOT lane on SR-167 in Seattle, Washington, should include a minimum toll, which is higher than the transit fare in the same corridor. Environmental justice is an important aspect of social equity and should be given a fair con- sideration. During the planning and design phase for dedicating lanes for CAVs, it is important to ensure that the project does not have a disproportionately high and adverse human health and environmental effect, which often has a greater or more direct impact on minority and lower- income populations. It is also important to prevent denial of, reduction in, or significant delay in the receipt of DL benefits by minority and low-income populations. Similar to HOT lanes, the CAV DL should be carefully evaluated in all categories of equity, as the mobility of the CAV DL is expected to be higher than that of the GPL due to the inherent advantages of CAV tech- nologies (e.g., V2V communication, V2I communication, CACC, self-driving). No policy, not even a well-designed and carefully considered one, will impact all groups equally. The operator
Categories of Benefits and Disbenefits to Stakeholders 19 should strive to attain the most equitable and feasible distribution while factoring in community acceptability and the overall transportation system impacts. 2.3.4.3 Public Outreach Public acceptance is vital for the sustainability of any DL strategy. Cases have shown that a near-empty lane can adversely affect public support. Public outreach about âempty-lane syn- dromeâ is necessary to ensure understanding that the DL actually moves more people through the corridor despite carrying fewer vehicles. Agencies can perform stakeholder engagement activities prior to the planning stage to assess the level and outlets of public outreach to help develop project goals and objectives. Furthermore, outreach efforts from government agencies (e.g., U.S.DOT, state DOTs) need to be actively performed through various types of activities (e.g., webinars, educational outreach, public hearings, field demonstrations) to achieve success- ful deployment of CAV and ITS technologies.
