Microtransit or General Public Demand–Response Transit Services: State of the Practice (2019)

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.

13 Survey Methodology The purpose of any TCRP synthesis is to summarize the current state of the practice within the transit industry, usually employing a survey of public transit agencies that provides infor- mation and insights into their experiences. The topics of synthesis reports often are common to almost all transit agencies, and the principal investigator can randomly select a subset of agencies to survey with the expectation that they all have some experience in the subject of the synthesis. In the case of this synthesis, the development of efficient and dynamic general public demand– response services is relatively recent and of great interest among transit agencies. There is no list of the transit agencies, however, that engage in dynamic general public DRT/microtransit. To identify the transit agencies that do engage in dynamic general public DRT/microtransit, this project relied on information from a variety of sources: • TCRP SB-30 project panel members • The Transportation Research International Documentation (TRID) database and the Transit Cooperative Research Program • The Leadership APTA program of the American Public Transportation Association • The Community Transportation Association of America • Broad internet searches through Google, Yahoo, and Bing • State Transit Association Directors • Passenger Transport (biweekly news magazine of APTA) • A variety of e-zines and newsletters produced by Planetizen, SmartCities, Metro Express, CitiLab, TransitNews.net, or ITS America. The preceding resources helped identify a relatively small number of agencies that engage in general public DRT that is dynamic enough to provide service within 2 hours of a passenger’s request. However, the most successful method of finding such agencies was through sending an e-mail and a survey instrument to the 200+ general managers, CEOs, and directors of transit sys- tems that are members of APTA. Their contact information was available through the electronic membership directory on the APTA website. The e-mail and survey sent to the transit agencies are in Appendix A. Because of this discovery process, 22 transit agencies fully participated. Five of those 22 agencies are in the process of planning for or procuring demand–response service but were still able to provide information of value to the project, including their request for proposals in one case. Fortunately, the information received from those 22 agencies is representative of a variety of meth- ods of providing dynamic demand–response service. The 22 agencies represent a good cross section of system sizes with six small (100 or fewer buses), 11 medium (101 to 499 buses), and five large (500 or more buses) transit agencies. The geographic distribution tilts toward the western half of the C H A P T E R 3 Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/ Microtransit Service

14 Microtransit or General Public Demand–Response Transit Services: State of the Practice United States as shown in Figure 1. The information below the map indicates agency name, when the general public demand–response service was started and ended, or if the service is ongoing. Table 1 shows the basic characteristics of these transit agencies and the services they provide. Note that every transit agency in the table, with the exception of the Greater Dayton Regional Transit Authority as described, is providing general public DRT only in portions of their service area and only during select hours. Reasons for Providing General Public DRT/Microtransit Service Operational Efficiency and Reduced Costs The frequently cited reason among survey respondents for establishing general public DRT service is to retain or provide service in areas of low demand or during times of low demand in the most efficient manner. As Mark Finnicum, chief operating officer of the Stark Agency Name Year Started Status Hillsborough Area Regional Transit Authority (HART) 2016 Ended 2018 Kansas City Area Transportation Authority (KCATA) 2014 Ended 2015 Salem–Keizer Transit (Cherriots) 2015 Ended 2017 Alameda–Contra Costa Transit District (AC Transit) 2016 Ongoing Central Florida Regional Transportation Authority (LYNX) 2008 Ongoing Dallas Area Rapid Transit (DART) 2006 Ongoing Denver Regional Transportation District (RTD) 2004 Ongoing Des Moines Area Regional Transit Authority (DART) 2002 Ongoing Greater Dayton Regional Transit Authority (GDRTA) 2017 Ongoing Kitsap Transit 2015 Ongoing Houston METRO 2015 Ongoing Monterey–Salinas Transit (MTS) 2010 Ongoing Napa Valley Transportation Authority (NVTA) 2015 Ongoing North County Transit District (NCTD) 2011 Ongoing Sacramento Regional Transit District (SacRT) 2018 Ongoing San Joaquin Regional Transit District (SJRTD) 2009 Ongoing Transit District of Utah (TDU) 2008 Ongoing Gwinnett County Transit (GCT) Planning Los Angeles County Metropolitan Transportation Authority (LA Metro) Planning Maryland Transit Administration Planning Regional Transportation Commission of Southern Nevada (RTC) Planning VIA Metropolitan Transit (VIA) Planning Figure 1. Public transit agencies that responded to the project survey.

Transit Agency Number of fixed route buses Population of service area Total number of annual unlinked trips Characteristics of general public demand–response service provided Alameda –Contra Costa Transit District (AC Transit) 500 1,425,000 54,575,655 • point deviation • call-a-ride service • first mile/last mile Central Florida Regional Transportation Authority (LYNX) 312 2,400,000 27,387,837 • call-a-ride service • first mile/last mile Salem Area Mass Transit District (Cherriots) 53 236,632 3,637,866 • call-a-ride service • point deviation • first mile/last mile Dallas Area Rapid Transit (DART) 533 2,380,530 66,799,954 • fixed route deviation • call-a-ride service • first mile/last mile Greater Dayton Regional Transit Authority (GDRTA) 124 559,062 9,973,237 • some service for the entire service area • some only in portions of the service area • some during all hours • some only during select hours • point deviation • call-a-ride service • first mile/last mile Denver Regional Transportation District (RTD) 873 2,920,000 103,340,797 • point deviation • call-a-ride service • first mile/last mile Des Moines Area Regional Transit Authority (DART) 113 374,910 4,775,768 • fixed route deviation • call-a-ride service • first mile/last mile Gwinnett County Transit (GCT) 63 907,135 1,496,448 • in the planning phase Hillsborough Area Regional Transit Authority (HART) 162 875,598 14,523,002 • call-a-ride service • first mile/last mile service Metropolitan Transit Authority of Harris County (Houston METRO) 937 4,298,000 89,970,895 • call-a-ride service • first mile/last mile service Kansas City Area Transportation Authority (KCATA) 179 788,748 14,220,399 • call-a-ride service Kitsap Transit 90 254,183 3,549,994 • call-a-ride service Los Angeles County Metropolitan Transportation Authority (LA Metro) 1,935 8,626,817 432,985,182 • in the RFP phase Maryland Transit Administration 907 7,811,145 110,727,565 • in the planning stages Monterey –Salinas Transit (MST) 75 433,898 4,406,784 • call-a-ride service Napa Valley Transportation Authority (NVTA or Vine Transit) 24 138,000 1,214,969 • call-a-ride service • contemplating first mile/last mile service North County Transit District (NCTD) 137 849,420 12,005,664 • point deviation • fixed route deviation Regional Transportation Commission of Southern Nevada (RTC) 341 2,008,655 67,346,272 • • in the planning phase for call-a-ride service Sacramento Regional Transit District (SacRT) 162 1,031,946 24,330,247 call-a-ride service San Joaquin Regional Transit District (SJRTD) 70 753,226 4,047,559 • call-a-ride service • fixed route deviation Transit District of Utah (formerly the Utah Transit Authority) 440 1,883,504 45,521,914 • fixed route deviation VIA Metropolitan Transit (VIA) 378 1,825,502 39,363,491 • in the RFP phase Note: RFP = request for proposal. Source: FY 2016 National Transit Database, Federal Transit Administration. Table 1. Characteristics of transit agencies that responded to the project survey.

16 Microtransit or General Public Demand–Response Transit Services: State of the Practice Area Regional Transit Authority in Ohio, who completed the survey for this project stated, “The costs of running fixed routes that aren’t producing to the level that make them self-sufficient isn’t good business. On the other hand, eliminating transit services to those riders who live on a route that isn’t self-sufficient isn’t a viable option either.” When addressing areas that do not generate the typical level of demand needed to support fixed route services, transit agencies are between the classic rock and a hard place. Transit agen- cies make every attempt to operate efficiently to realize taxpayer expectations. They are often guided by service standards that establish and require certain minimum levels of ridership to maintain transit service. A significant number of communities within transit agencies’ service areas have received no service at all due to the inability of the areas to generate ridership that meets adopted service standards. General public DRT is typically less expensive to provide than fixed route service and therefore an appealing alternative for transit agencies to consider. Jurisdictional Equity Regardless of ridership potential, areas of low transit demand provide some of the tax sup- port transit agencies need to operate. Consequently, the perceived need to address the issue of political/jurisdictional equity has often been a factor in the decision to provide general public demand–response transit. This type of service can demonstrate that communities that provide tax support for transit can receive some form of service in return. General public DRT/microtransit is a way of more cost-effectively addressing this issue, whether the service is provided where no transit service exists or to replace fixed route service that would be eliminated due to low ridership performance. Expanding Economic Opportunity Other equity issues exist that general public DRT can help address. Transit agencies fully understand the importance of the roles they play in providing mobility to lower income people. Houston METRO managers noted via Jim Archer, director of service planning and evaluation, who completed the survey that “There is a genuine concern for those who are socio-economically disadvantaged and a desire to assist them with access to opportunities,” including jobs, educa- tion, and social services. Some of these opportunities might be in newly developed areas not yet served by transit. Lower income people might live outside the areas currently provided with transit service and lack access to the local system. While fixed route service might be too expen- sive to provide, lower cost general public DRT could be the most cost-effective way to provide some level of mobility for those who need it in these circumstances. Continuing Service for the Elderly and for Those with Disabilities The Transit District of Utah (TDU, formerly the Utah Transit Authority) addressed another equity issue when it implemented fixed route deviation services to meet the needs of ADA paratransit service recipients. In 2008, the Great Recession reduced the tax revenue supporting TDU, causing the agency to move to eliminate some of their lowest performing fixed routes. This would have caused many ADA-eligible customers to lose their access to paratransit service since there would be no more three-quarter-mile buffer zone around an eliminated route. Replacing many of the former fixed routes with route deviation service was a solution that allowed the agency to provide mobility options for those ADA paratransit riders while also maintaining service that resembled fixed route that other riders could use as well. Because TDU’s Flex ser- vice provides route deviations (two per trip) on an equal opportunity basis, the service fulfills Federal Transit Administration (FTA) mobility requirements for ADA riders without incurring a separate paratransit obligation.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 17 In another example of concern for a particular segment of their population, the Greater Dayton Regional Transit Authority is planning a major revision to its transit, paratransit, and general public DRT services. This revision would “blur the lines” between modes to ensure that the growing elderly population in their service area has sufficient access to the mobility they need and to age in place. AC Transit viewed demand–response service as the most appropriate way to serve a major senior center in an area lacking transit service, while Vine Transit in Napa Valley, California, implemented call-a-ride service for a major veterans home in the Yountsville community. Miscellaneous Reasons for Providing General Public DRT/Microtransit Service Transit agencies either have implemented or are exploring establishing general public DRT for several reasons. Cherriots in Oregon instituted DRT in a community with disconnected streets, insufficient infrastructure in terms of sidewalks, and hilly terrain that made operating large buses challenging and walking to fixed route bus stops difficult and sometimes dangerous. The Regional Transportation Commission of Southern Nevada (Las Vegas), the Hillsborough Area Regional Transit Authority (Hillsborough County, Florida), the Sacramento Regional Transit District, and many others view demand–response service as complementing their regional systems as they serve to feed to, and distribute from, express routes, rail lines, or the rest of their regional system. The service helps to address the first mile/last mile challenge that has faced transit agencies for years. Dallas Area Rapid Transit (DART) also initiated On Call service for the general public to replace inefficient, declining bus service in small zones around certain rail stations and transit centers. Similar to Cherriots’ experience, many areas outside Dallas had cul-de-sacs, walled communities, or other conditions making fixed route service inefficient, yet those areas also had senior citizens and some other minor demands for transit. DART reports that On Call DRT has proved to be less costly than adding a network of bus service. The Denver RTD provides Call-n-Ride service in areas with dispersed travel patterns that make fixed route transit impractical. Transit agencies realize that “one size does not fit all” and a growing number appreciate the flexibility general public DRT/microtransit services can provide. On the one hand, the services can replace fixed route service that is performing well below service standards. On the other hand, the services can be in areas that have been previously unserved to test the market for tran- sit service. As a transit agency gains experience and knowledge of the patterns of demand, the agency can adjust the service, potentially resulting in a service going from Call-a-Ride, to fixed route deviation, to fixed route. This happened in the Denver area (5). The most common theme is that general public DRT service is becoming another tool in the toolbox that allows transit agencies to match the appropriate level of supply to the existing level of demand. In general, it is a type of service to consider when fixed route services simply do not make good sense. While there is often an attempt to have private microtransit services in major cities to secure as much farebox revenue as possible, the DRT/microtransit services provided by the public sector are typically provided in underserved, low-density areas or at times of the day when it is expected to operate more efficiently and effectively than fixed route service. There are two final reasons that more transit agencies reported they have implemented or are considering implementing general public DRT/microtransit. The first reason is due to chang- ing customer expectations regarding demand–response transportation in general. While public transit does not duplicate or supplant taxis or TNC services, there is a realization that today’s customers now experience and expect more responsive transportation service. Private sector attempts at providing microtransit service have found themselves squeezed between the high

18 Microtransit or General Public Demand–Response Transit Services: State of the Practice cost of complying with regulations and providing a service at a fare that people can afford. Per- haps general public DRT/microtransit can satisfy a market attracted to more customized service that is affordable while ensuring safe vehicles and qualified drivers (17). There might be a sense of urgency among some transit agencies concerned that private microtransit could grow, causing public transportation ridership to further decline. The second reason for implementing such service is that they can. Increasingly affordable technology has provided the opportunity for transit agencies to engage effectively in demand– response service that is truly dynamic and not reliant on requiring customers to call for a ride a day in advance. More detail on the technology platforms used is provided in the section on technology used to make DRT efficient and customer friendly. Planning and Design of General Public DRT/Microtransit Service Through the development of comprehensive transit development plans, transit agencies are able to identify the portions of their service area that are likely to support fixed route transit and that are not likely to support fixed route transit. These comprehensive transit development plans take into account a great number of factors, including population and employment densities, rates of vehicle ownership, household incomes, and commuting patterns, among many other things. With this information in hand and with knowledge of the historical performance of transit services in the area, transit planners can design the foundation for a fixed route system. It is not difficult for transit planners to identify the fixed routes that have underperformed. While every transit agency will set its own performance standards, routes that carry less than half of the system average in terms of passengers per hour will usually be subject to careful review to determine if there are steps that can be taken to improve their ridership. When routes perform noticeably worse than that, there may be discussion of whether fixed route service should be discontinued (20). In times past, this meant either continuing an inefficient fixed route, replac- ing it with even lower frequency bare-bones service, or discontinuing service completely. None of these options is attractive. Areas served by such fixed routes are potentially good candidates for general public DRT service. There are also low-density areas within a transit agency’s jurisdiction that might have never had transit service, but there have been requests for service. Circumstances might have changed that warrant a consideration to provide service in such areas. For instance, new development might have occurred in a previously sparsely developed area or a neighborhood’s socioeconomic characteristics might have changed to include more people of lower income. It is more difficult to predict exactly where a fixed route should be placed within these areas since there is no history of performance, and overall or geographically specific demand cannot be confidently forecast. In either case, general public DRT is becoming an increasingly attractive option that might serve a community’s needs better than fixed route transit. General public DRT is sometimes referred to by different names such as microtransit, on-demand transit, on call, call-and-ride, dial-a-ride, or Flex service, but all the names generally fit into the following paradigms of service: • Many to many/many to few, on-demand, community-based • Feeder to transit network through scheduled connections • Point deviation (Flex route): DRT with dynamically or regularly scheduled checkpoints • Route deviation (Flex route): fixed route with regularly or dynamically scheduled, off-route pickups or drop-offs (5) DRT provided through the many-to-many paradigm can be door to door, curb to curb, or available at a limited number of strategic stops in the community. This type of demand–responsive

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 19 service provides more “coverage efficiency,” by serving dispersed origins and destinations at rea- sonable costs without unnecessary detours to stops where there might or might not be demand for service on a given vehicle trip (21). While providing services at strategic stops requires a bit more walking by passengers, it helps the bus serve more of the area in a zone in a set amount of time. While not being as predictable as a fixed route service, it makes the service more accessible to more people within the zone. The more structured DRT becomes, the more likely it is to increase productivity in terms of passengers per mile or per vehicle service hour, assuming there is a market to support the ser- vice within the DRT zone. The feeder to transit network paradigm typically features the DRT vehicle connecting with a major transit bus or rail station and addresses the first mile/last mile challenge faced by many passengers, particularly commuters. DRT service operating in feeder service mode is governed by a cycle time, which is the round-trip time, including layover for the vehicle to return to its cycle point/checkpoint (5). The cycle time determines the bound- ary limits or size of the DRT service area since the minibus can only cover so much distance before needing to be back to the cycle point where it typically connects with regional bus or rail service. Additional vehicles placed into service within the zone allow for a larger service area or shorter headway or ride time (7) but also increase the expense to the transit agency. Deviation services operate with checkpoints as well. Point deviation services are usually the best choice when there are major generators of trips (e.g., schools, shopping centers, or office parks) within the zone while the rest of the demand is spread out. Buses can be at those major generators at scheduled times and respond to individual requests for service the remainder of their cycle time before returning to their primary terminal point. A minimal amount of DRT service can be provided at first to minimize risk but can be increased as patterns of demand become better known, which has been done by the Denver RTD. The paradigm can be modified over time from many to many to point deviation, to route deviation, to feeder service, or to a combination of all of them. The flexibility of general public DRT allows the community to have local circulator service as well as first mile/last mile service that connects to the regional system, although the availability of either might fluctuate through- out the day and particularly during rush hours. Within the context of microtransit, circulator service focuses on providing trips on-demand within a designated zone with many origins and many destinations within the zone. Once the candidate communities for general public DRT have been identified, focused plan- ning takes place. The specific type of DRT to be provided will be influenced by factors such as population and employment density, socioeconomic characteristics and travel behavior of the population, the location of traffic generators, the service area size, street layout and connectivity, and transit network connections. The terrain and infrastructure within a neighborhood might also affect the type of service selected. For instance, the DRT zone selected by Houston METRO had a series of apartments and schools with limited sidewalks and lighting, a non-contiguous street network, and deep open drainage ditches alongside the roads. These conditions made walking to a fixed route bus stop difficult and unsafe in some places and highlighted the need for customized demand–responsive service. The size of general public DRT zones is as flexible as the service itself. Survey respondents reported zones ranging from 1.1 square miles to 30 square miles. For simplicity’s sake, some DRT zones coincide with political boundaries. The size of the zone will depend on the type of DRT proposed, the level of demand expected from the zone, the known or anticipated markets within the zone, and the desired cycle times. The size of the zone will also depend on the num- ber of vehicles the transit agency is willing to place into service and the capability of the DRT software platform to accommodate and satisfy multiple real-time trip requests made within a set timeframe. Before launching a Call-n-Ride service, Denver RTD researched travel patterns,

20 Microtransit or General Public Demand–Response Transit Services: State of the Practice trip generators, and travel speeds to appropriately size the service area and designate checkpoints (i.e., locations where the vehicle is available at set times) and zones (5). Survey respondents noted how they have attempted to learn about the design of general public DRT from peers and any studies or reports they can find where such service has been imple- mented. One agency even noted that given how recently many of these services were developed and the dearth of standards for such service, transit agencies seem to be making it up as they go. However, service design is not an exercise of taking shots in the dark. Designing general public DRT usually requires a high level of involvement with the communities to be served. Since the service is different from what people are familiar with and there are many options as noted pre- viously, there is a need to work closely with the community to inform them of the differences and to receive their input on what would be most useful. The effort to receive input has involved surveys (including some that are web-based), focus groups, public hearings, workshops with key community stakeholders, information booths at community events, and community meetings at transfer centers and rail stations. The effort has also involved meetings and workshops with key community stakeholders, including area agencies, non-profit organizations, government staff, businesses, hospitals, entertainment centers, schools, training centers, colleges, and other likely attractors or generators of trips. Public support for general public DRT as a replacement for poor-performing fixed routes is not always a given. Houston METRO originally proposed the creation of five Flex zones to replace fixed route service. However, following a series of general and area-specific meetings with citizens, community leaders, elected officials, and stakeholders, it became apparent that there was not general support in all five proposed zones. The most significant problem encoun- tered was the perception that the proposed general public DRT service would be operated iden- tically to the pre-existing paratransit demand–response service. This perception was an issue because the pre-existing paratransit demand–response service was not held in high regard in several areas where the new DRT service was proposed. Consequently, only one Flex zone for general public DRT was established in the Houston area. The North County Transit District in southern California anticipated that not all proposed Flex routes would be well received. On the basis of customer surveys, the District removed two proposed Flex routes and retained fixed route service while instituting route deviation DRT in three other zones. These examples dem- onstrate the need to be familiar with the market and aware of a community’s needs and desires before finalizing the appropriate service. In addition to using their own staff or traditional transit consultants, a number of agencies are procuring services from more newly formed technology companies that specialize in the development of software to support general public DRT/microtransit. These companies help to estimate the demand for service, suggest the size of the area to be served, and run simulations of various scenarios to present alternatives for service delivery. They develop the software to enable dynamic scheduling or reservations and communications necessary for the DRT service and sometimes provide the hardware as well. Because these companies specialize in this type of service, they can help inexperienced local transit agencies understand the issues, problems, and potential associated with general public DRT/microtransit before full deployment. However, they will need the transit agency staff’s local knowledge of the area to be served to develop the realistic estimates for demand. Table 2, which was developed by a microtransit technology company, illustrates the options to consider in terms of numbers of trips and vehicles to provide and what the likely results would be. The table provides answers to questions such as how long will the ride time and wait time be, how many vehicles would have to be added to serve a larger number of passengers, how costs would be affected, and what the estimated riders per hour would be with different numbers of vehicles.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 21 The growing interest in and potential of general public DRT/microtransit have influenced the broader planning and design of entire transit systems. AC Transit created a task force that drafted a plan to increase service frequency to 15 minutes on key arterials and replace low-frequency routes with general public demand–response zones. The intent is to increase ridership by providing fre- quent service on key fixed routes through the savings associated with replacing poorly perform- ing fixed routes with general public DRT. As noted in the Eno Foundation’s report UpRouted: Exploring Microtransit in the United States, “AC Transit identified specific use-cases to test, along with an overall vision of what the future of Flex would look like in the region. Headways for AC Transit’s preexisting services were 45 to 60 minutes, yielding a service that was not useful for most customers. Staff hypothesized that if they could bolster service on the core high frequency routes and deploy flexible, on-demand options across existing low frequency feeder routes to create a network effect, they would be able to grow overall ridership, and in a cost neutral way.” (12) Although the response to AC Transit’s general public DRT has been generally favorable, the ridership on the DRT routes has not performed as well as the ridership on the fixed routes they replaced during off-peak hours. It is too early to tell if the minor changes associated with the agency’s institution of DRT in two service areas have resulted in increased ridership on other fixed routes. However, the agency still intends to shift its overall system plan to a 70% ridership–30% coverage DRT split, using the same spending level as today. The agency believes that demand–response service is the only way they can invest in frequency while main- taining coverage obligations. A number of transit agencies reported that general public DRT has allowed them to establish coverage zones that are not subject to the same performance guidelines as fixed route service. Those transit agencies do not expect or require ridership levels to be high or near their system’s average. However, the agencies appreciate that general public DRT/microtransit, though it carries far fewer passengers per mile or hour than fixed route transit, allows people within greater portions of their service area to have some form of access to the regional transit network through a service that is less costly and entirely appropriate from a political and socioeconomic equity perspective. Marketing of General Public DRT/Microtransit Service Transit systems have long used a variety of techniques to inform the public of their services, and most of the traditional techniques were cited to help promote general public DRT. As the previous section of the report noted, marketing a service that is new and unfamiliar to people presents new challenges. The major issue facing the Greater Dayton Regional Transit Authority during their planning process was ensuring that their customers understood the new alternative service being considered. Staff worked diligently to explain to customers that while the proposed Number of Trips Number of Vehicles Hours of Service Rides per hour per Vehicle Average Wait Time (minutes) Average Ride Time (minutes) Cost per Trip ($) Ride Pooling % 50 2 11 2.3 2 17 28 25 100 2 11 4.5 10 24 14 41 100 4 11 2.3 2 15 28 20 100 6 11 1.5 0 15 43 15 300 6 11 4.5 8 23 14 40 300 10 11 2.7 2 16 24 25 Note: Ride pooling represents the percent reduction in number of vehicle stops due to pooling of rides. Cost per trip is based on $65 per vehicle hour. Source: TransLoc, Inc. (22). Table 2. Performance parameters from a recent microtransit simulation.

22 Microtransit or General Public Demand–Response Transit Services: State of the Practice service was different from their current service, it would be better. Their outreach and educa- tion efforts had a positive impact that resulted in a service that is still evolving but has been well received. The Napa Valley Transportation Authority also believed that most people would have a difficult time coping with general public DRT at first since change is something many, if not most, people have a difficult time easily accepting. This problem was realized but was short lived as residents of each community began to see how the inherent flexibility and the opportunity to have curb-to-curb service enhanced convenience and improved their mobility. Cherriots emphasized aspects of their DRT service not normally associated with transit marketing. They advised citizens of the area to be served that they would no longer see big buses driving past their homes with few passengers in them and that the smaller minibuses would only come to a pickup or drop-off point when someone requested it. When there were no requests for rides, the bus waited at a nearby transit center. In those circumstances, the agency could make a good case that less might be more. Many transit agencies approach the initial provision of general public DRT as a pilot project to last approximately 1 year or less, with a review of performance after that time to determine if it should continue, be modified, or ended. Approaching the provision of new service as a pilot or as an experiment brings some added interest to the service, and this approach might also help those who are skeptical to more readily give the service a chance. The literature search and the survey results show that not every pilot project has been renewed or continued [e.g., Santa Clara Valley Transportation Authority, Kansas City Area Transportation Authority, the Salem Area Mass Transit District (Cherriots), and Hillsborough Area Regional Transit Authority all discontinued their initial efforts]. Gwinnett County Transit in Georgia reported that it is using TransLoc, a provider of DRT technology platforms, to assist them with promoting their upcoming DRT service. The transit agency, with only five full-time employees and a limited budget, finds this marketing assistance almost as valuable as the simulations that the company will provide. General public DRT/microtransit may operate differently from fixed route, but it is still public transit, and, in that regard, familiar marketing methods were reported that were used to promote the service. Direct mail marketing can be sent to virtually every residence and business in a DRT zone and is used often, because the entire market is in a well-defined and limited geographic area. More targeted follow-up efforts can be made with entities likely to generate or attract the most trips such as colleges, hospitals, schools, apartment complexes, industrial complexes, office parks, and entertainment venues. Responding agencies noted that public outreach efforts occur before, during, and after implementation of the service, because adjustments are anticipated due to changing demand as people become more familiar with the service. The following list represents the types of marketing actions survey respondents reported using to promote their general public DRT/microtransit service: • Distributing brochures, flyers, seat drops, and car cards. • Placing ads at rail stations and transit centers that serve as DRT terminals. • Putting ads on bus exteriors and bus shelters. • Advertising on digital media. • Placing information on the transit agency website, including a new carousel image, social media, and eNews. • Using bilingual brand ambassadors at rail stations and on buses affected by the change to help passengers understand and register for the service. • Placing at-stop signage, including pole case inserts and Flex service signs. • Distributing copies of Guides to Proposed Mobility Plan Changes at outreach sessions and on transit vehicles.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 23 • Placing “Take-ones” on all fixed route buses identifying changes and locations of public open houses where people can provide feedback. • Conducting employee meetings at the businesses and housing developments within the DRT On-Call Zone. • Purchasing radio spots and television commercial time on English and Spanish channels. • Providing information through the transit agency’s customer service center. • Conducting direct outreach to political jurisdictions, planning commissions, government agencies, and human services agencies. • Holding public meetings, public hearings, and making presentations at community centers and public spaces. • Advertising and placing a featured story in the local papers. • Mailing information brochures to large senior living facilities and mobile home parks. • Holding press conferences, issuing press releases, and arranging television news coverage. • Mailing introductory notices along with free ride coupons to residents living within 1.5 miles of newly established Flex Routes. • Presenting the general public DRT/microtransit vehicles at an organized event with public officials in attendance drawing media coverage. • Having ribbon-cutting ceremonies at the start of the new service. How General Public DRT/Microtransit Is Integrated into an Agency’s Family of Services Branding Virtually all of the transit agencies that responded to the survey indicated that their DRT service is (or was until eliminated) included as an integral part of their suite of services in all their literature and websites. There are often distinct identifying brands created for the different DRT services such as • “NeighborLink” is the on-demand service provided by the Central Florida Regional Trans- portation Authority (LYNX) in the suburbs surrounding Orlando. • “Call-n-Ride” is the dial-a-ride service provided in multiple zones around Denver. • “The Connector” is the microtransit service in West Salem, Oregon, that the Cherriots provided. • “On Call” is the on-demand service provided in multiple zones in Dallas, which prioritizes providing first mile/last mile connections to or from rail stations. Monterey–Salinas Transit (MST) uses the same name for dial-a-ride service provided in cities outside their fixed route services. • “RTA Connect On-Demand” is the name for the family of DRT services provided in the greater Dayton, Ohio, service area. • “SmaRT Ride” is the microtransit service provided in suburbs in the service area of the Sacramento Regional Transit District. • “RTD GO!” is the dial-a-ride service provided in more rural areas surrounding Stockton, California. • “The Hopper” is the route deviation service provided in Stockton by the SJRTD. By far, the most popular name given to DRT services offered by survey respondents is “Flex.” Houston METRO, AC Transit, the Transit District of Utah, and the North County Transit Dis- trict use this name, and HART used this name before its service was discontinued. Using such names to help identify the DRT service is no different from transit systems providing a distinct brand name for bus rapid transit service, light rail service, paratransit service, or express service. Although the DRT service may be given a particular brand name to draw attention to it and

24 Microtransit or General Public Demand–Response Transit Services: State of the Practice distinguish it from other modes, the service is still regarded as an integral part of the transit agency’s services, whether provided with their own personnel and equipment or through a con- tracted service. The vehicles used might have a distinct branding or logo, but they are also clearly marked as being part of the transit agency that is funding the service. Fares The majority of survey respondents reported that the fare policies governing general public DRT/microtransit are the same as those that govern the rest of their fixed route transit services. In those cases, the same fare structure and fare media are used for both modes. The same methods of payment are typically used. When the same fare policies govern the DRT service, transfers to or from DRT are treated the same as transfers between two fixed route buses. When there are no transfers between fixed routes, there are likewise no transfers to or from DRT services. The Greater Dayton RTA reported that they offer their On-Demand services free, as does the Napa Valley Transportation Authority in two of its Flex zones. Gwinnett County Transit offered free fares for the pilot Flex service they instituted in the summer of 2018, which initially serves only as a circulator with no connections to the county’s fixed route services. They believe this isolated pilot will give them time to figure out, before going to a full launch, how they will con- nect the fares for the Flex service with the rest of the regional smartcard network. Another reason they are offering fare-free service initially is that they are leasing minibuses and do not have spare fareboxes to place on these vehicles. Of course, offering free fares represents an additional marketing strategy to encourage people to try the new service. A few survey respondents reported that they charge different fares for their DRT services. HART charges a separate $1.00 fare for their Flex service and issues no transfers between the Flex service and their fixed route service. The agency recommends that passengers consider the pur- chase of an all-day pass for $4.00, which provides unlimited use of all HART services for the day. Similarly, NCTD issues no transfers between their FLEX service and other modes with the exception of those passengers with a Region Plus day pass ($12.00) or a one-zone Coaster pass ($120.00), which are valid on all NCTD services. Otherwise, passengers must pay a full fare when boarding FLEX service or transferring to other services. NCTD considers FLEX a premium, more custom- ized service, which is reflected in their fares of $5.00 for a one-way trip, considerably higher than the $1.75 fare charged for regular fixed route service. They also charge a $10.00 fare for one Flex route that is 19 miles long but note that a similar trip via a TNC would cost over $30.00. DART in Dallas also charged premium fares when they first started providing dial-a- ride services but found that fare structure to be unpopular and discontinued doing so. Passen- gers in Poughkeepsie, New York, pay $7.00 for recently instituted dynamic dial-a-ride services that only require a reservation to be made 30 minutes in advance instead of a day in advance. The Transit District of Utah, SJRTD, and DART in Des Moines charge passengers who request deviations on their Flex routes a different fare from those who board or alight along the fixed route. TDU passengers must pay an additional $1.25 to accommodate their deviation request, representing an additional 50% above the regular fixed route fare. SJRTD charges an additional $1.00, representing a 67% increase above their fixed route fare of $1.50. DART charges an addi- tional $1.75, representing a 100% increase over the regular fixed route fare. Integration with Other Transit Services Transit agencies regard their general public DRT service as simply another multimodal service offering. Customers can call one telephone number and visit just one website for information on each transit agency’s entire family of services. While the fares for DRT services sometimes differ from the rest of the network as previously noted, the same forms of payment are accepted.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 25 Though DRT Flex zones can offer point-to-point circulation services for trips within the zone, it is clear that the intention of the general public DRT services described in this report is to integrate with and complement the rest of the transit network in the area. The names given to some of the services, such as The Connector, RTA Connect On-Demand, and NeighborLink, convey this intent. As just one example among many, all of NCTD’s FLEX routes either begin or end at one of NCTD’s major transit centers, where customers may connect with fixed route, paratransit, or rail service. Additional fixed route bus connections are available along each of NCTD’s FLEX routes as well. Denver RTD’s senior manager of service development has more than 15 years of experience in designing and implementing general public DRT. He reminds those who are contemplating establishing DRT services that 70% of all transit trips are for work or school purposes and often require travel beyond the DRT zones (interview with Jeff Becker, March 30, 2018). First mile/last mile services allow people to start or complete their trips by using the general public DRT service in conjunction with the rest of the region’s transit network. As noted earlier, transit agencies embrace the purpose of providing mobility for people of all income status to enhance their access to opportunities throughout a region. With only one exception, all survey respondents indicated that their DRT services make connections with the regional transit network. Those connecting points are almost invariably the locations where the most boarding or alighting takes place within the DRT zone. These connecting points also offer one of the best opportunities for people to board a DRT service vehicle without needing to make a reservation. The hours of service might or might not coincide with the full span of service provided in the rest of the regional transit network. For instance, DRT zones in Dallas typically end service by 8:00 p.m., as do the on-demand services provided by LYNX. On the other hand, the Greater Dayton Regional Transit Authority makes service available on a 24/7 basis, though their experi- ence is that customers are only using the service during times when they can connect to fixed route services. HART initially provided DRT service 1 hour prior to and 1 hour after completion of the regular bus schedule to allow passengers to catch the first fixed route bus of the day and return home from the last bus of the day. Based on ridership experience, hours were shortened to start at 6:00 a.m. and end at 10:30 p.m., 7 days a week. DRT services in noncommuter markets normally provide service fewer hours of the day. For instance, MST’s On Call service in many lower density residential communities starts at 8:30 a.m. and ends before 6:00 p.m. In general, the hours of service established for DRT services will reflect the markets that are served and are adjustable based on the experiences the agencies gain and community feedback. Factors That Influence When to Contract for DRT Service The decision whether to contract for general public DRT can be determined at times by competitive procurement and at other times by public policy. The AC Transit Board, which is comprised of people elected directly (not appointed) to the board, has strong support from its operators’ union and generally does not contract any transit service. Beyond just election politics, the board members are concerned about the level of professionalism of contracted drivers and the increasing loss of jobs and income suffered by people in the gig economy. The Sacramento Regional Transit District also utilizes its own equipment and personnel for service in the Citrus Heights community, in which dial-a-ride service had been provided but required reservations to be made a day in advance. The agency wanted to start the microtransit pilot quickly and used the same vehicles and employees. The only operational aspect that was altered for the pilot was the use of new proprietary software that allowed dynamic reservations and scheduling.

26 Microtransit or General Public Demand–Response Transit Services: State of the Practice The Stark Area Regional Transit Authority in Canton, Ohio, which has contracted with TransLoc to start two microtransit pilot projects in the second quarter of 2019, will use its own employees. They expect the service to carry many of their ADA-eligible customers and do not want the care of those passengers to be outsourced to a third party if it can be avoided. Houston METRO started providing DRT through a pilot project in March 2015 by using only contracted buses, but the services were brought in house after METRO created a new position called “service driver,” which helped reduce the costs associated with providing DRT with their own personnel. The clear majority of survey respondents to the survey indicated that they provide general public DRT using private contractors secured through the competitive procurement process. A number of studies have shown that contracted transit services are usually less expensive (23, 24), but the reasons for contracting vary and are sometimes nuanced due to particular local circum- stances. DART in Dallas decided to contract for their on-demand service because it allowed them to operate with lower overall costs per hour. However, the agency provided the vehicles and the operating facilities for the contractor. In the future, DART might have the contractor purchase the vehicles if they are able to acquire vehicles more quickly. DART found that the contractor was more adept at providing increased or decreased service more quickly than would have been possible with DART employees operating under the collective bargaining agreement. This factor is important when providing a new type of service in areas where the level of demand is uncertain and might require adjustments to service levels quickly. The same contractor also provided DART’s ADA paratransit service, which was helpful not only because of the contrac- tor’s familiarity with the service area but also because paratransit service has many similar char- acteristics to general public DRT service. HART noted that existing collective bargaining agreement provisions would increase the cost of the service if provided with their own personnel. New service that bargaining agreements did not cover could be contracted. HART staff found the contracted service to have lower costs for labor, maintenance, and capital items. The agency’s RFP for the DRT service attracted three bidders. HART was pleased that the selected contractor had experience in DRT service, an excellent training program, and an app already created to allow for dynamic scheduling and reservations. NCTD has historically outsourced paratransit and demand–response services to reduce oper- ating expenses, and in 2010 outsourced all fixed route service as well because of reduced rev- enues suffered through the Great Recession and increasing costs. Outsourcing saved the agency enough money to prevent what would have otherwise required a 25% reduction in fixed route services (25). In late 2016, NCTD issued a combined fixed route, paratransit, and specialized transportation RFP, and the new contractor operating all three modes started on July 1, 2017. The Colorado state legislature has required RTD to contract a substantial portion of its services for many years. As a result, RTD contracts over half its bus service and all its ADA paratransit service, while providing the vehicles and fuel but not the garage, operators, or maintenance. The contract for RTD’s Call-n-Ride service operates the same way. The agency believes this model results in the greatest amount of competition and cost savings, while requiring all bids to include reasonable wage structures. This model also provides greater flexibility for geographic locations of operations centers and easier rightsizing of the supply of service. In a slightly different model, the Napa Valley Transportation Authority owns the assets (vehicles and facilities) for all their services, while the contractor provides the drivers and maintenance personnel. MST decided to contract for its DRT Dial-a-Ride service not only to lower costs but also due to the limited space available at its operating bases to accommodate additional vehicles. They were also favorably impressed with the contracting experiences reported by other agencies.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 27 LYNX started its NeighborLink on-demand service through its contract for paratransit ser- vices. After the implementation of multiple routes, the source was a separate solicitation and contract. A smartphone app released in 2017 reduced the call volume and number of reserva- tions handled by dispatchers, allowing the dispatch functions to be brought in house. Providing a cost-efficient solution and greater connectivity to the fixed route system were the two most important factors in bringing the dispatch function in house. The California Public Utilities Code mandates that the SJRTD contract out all transit services outside the Stockton Metropolitan Area to realize lower costs for providing services in rural areas. The Dial-a-Ride DRT service in rural areas is provided through a contractor while the DRT route deviation services within Stockton are provided by RTD personnel and equipment. At the Transit District of Utah, whoever provides paratransit service in an area also provides the DRT route deviation service. TDU personnel provide the DRT service in the largest geographic portion of their service area, but in areas where they contract out paratransit service, the contractor provides demand–response route deviation services as well. TDU staff schedules all paratransit and route deviations; builds the route structures; and develops, delivers, or approves driver training. The transit agency administers all the administrative functions, including customer care as well. Contractors essentially provide the driver, radio dispatcher, and the supervisor/training staff in accordance with the route structure, daily demand (for paratransit), and TDU’s public sched- ules for the route deviation service. To summarize, the various reasons for a transit agency to consider contracting for general public DRT service include • Where state law or local decision makers require contracting for all or a portion of their services. • Where state or local law allows pilot projects to be undertaken without a need to seek competitive bids. • When there are clear cost savings (due primarily to lower labor costs). • When the contractors have substantial experience in providing demand–response service. • When a private contractor provides greater ability to modify service on short notice than what would be possible under the transit agency’s collective bargaining agreement. • When a private contractor already provides ADA demand–response service for the transit agency. • When a transit agency has insufficient space to accommodate the vehicles for the DRT service. Procurement Processes and Key Provisions Within Contracts for DRT Service Establishing general public DRT/microtransit as a pilot project provided a number of survey respondents with the ability to establish their services without going through lengthy competi- tive bid processes. Cherriots took advantage of the fact that Oregon procurement laws permit- ted sole source contracts for pilot projects. Since no federal funds were used for the project, the transit agency was able to secure the services of a technology company to help develop the communications/scheduling/automated reservations system needed to implement on-demand microtransit service in the hills of West Salem. Similarly, the Greater Dayton RTA has established their Connect On-Demand service as a 2-year pilot project. Any eligible provider was invited to participate in the delivery of DRT services simply by signing the RTA’s general services agreement. NVTA also utilized its existing contract for all its fixed route and paratransit services without specifically seeking separate com- petitive bids for their on-demand services. The new DRT service (call-a-ride) is not that different

28 Microtransit or General Public Demand–Response Transit Services: State of the Practice from a previous service (fixed route deviation) they had in place, and the existing contract had provisions for a number of liquidated damages for unsatisfactory performance. They were also confident in the experience their contractor had with providing general public DRT. All the other agencies reported seeking competitive bids for their general public DRT ser- vices and receiving multiple proposals for providing the services, indicating the healthy inter- est there is among private contractors to provide such services. NCTD and MST included the provision of DRT as part of a more comprehensive request for multiple services, including paratransit and fixed route. Other agencies, including DART in Dallas, HART, and LYNX, issued requests for proposals specifically for their DRT services. DART provides the vehicles and operating garages as a technique for increasing the number of potential bidders. Their contracts include penalties for failing to meet standards for on-time performance and for not answering phones quickly. The contract did not include any incentives. There were also standards for safety. RTD wants to maximize competition when issuing RFPs for their DRT services. One way they encourage more competition is by not using low bids as their exclusive determinant of awarding contracts. RTD also gives great weight to the experience of the company and to the staff dedicated to the service. They believe this allows smaller companies to be competitive. RTD also wants to know the proposed salaries for contracted employees to ensure that the contractor will be paying realistic (not too low) salaries to help avoid turnover. The regional authority operating the public transit agency also carefully reviews the proposals to ensure staff- ing for the contract is adequate, and there are clear and specific descriptions of the services the contractor will provide. Additional Call-n-Rides can be assigned under their contracts at a variable rate. The fixed rate can be negotiated if service levels are increased or decreased by 15% of annual hours, because consideration of staffing levels may vary depending on the number of Call-n-Ride peak vehicles. Depending on additional assignments, additional staff such as supervisors, operations specialists, and extra board drivers may need to be assigned. Assign- ments are based on the geographic location of the Call-n-Ride contractor and on the quality of service (e.g., on-time performance or missed runs) that the contractor has provided in the past. MST’s key provisions for demand–response service include the contractor making every rea- sonable effort to pick up prescheduled passengers no more than 5 minutes before or after their scheduled pickup time. The contractor must make every reasonable effort to accommodate all trip requests and ensure that all passengers have been counted or tallied in the GFI-automated fare collection system. GFI is a manufacturer that supplies fare collection equipment to the public transportation industry. MST receives all customer service complaints or compliments regarding the DRT service, which helps MST to gauge the public’s overall satisfaction with the contractor’s service. NCTD specifies the requirements for contractors to operate FLEX services, including the type of vehicles used, descriptions of each route, the reservation call center, fare collection, schedules, and adherence to policies regarding on-time pickups and abandoned trips. If the contractor does not reach certain thresholds per the contract in on-time performance, miles between mechanical failures, miles between accidents, percentage of service operated, and complaints per passenger, the contractor is assessed a financial penalty. The key performance indicators for contractors providing point deviation services for LYNX are missed trips, on-time performance, and meeting connections with fixed route services. HART conducted quality assurance checks, including the use of random secret shoppers who rode the service provided by the contractor. HART reported conducting Level 2 background checks on contract employees as well as drug and alcohol testing. In addition, the agency con- ducts vehicle inspections and monitors contractor performance on a monthly basis.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 29 AC Transit provides Flex service with its own personnel and vehicles but issued a competi- tive RFP for a software as a service scheduling and ride-matching software component. The agency included a task for customization to AC Transit’s needs but did not ask to own the modifications, as some agencies do. The modifications included tasks for system and user training, cloud hosting, system monitoring and maintenance, and technical support. The technical support was key and, as the agency reviews the process it went through, they believe they should have specified more detail in the contract about response times and who handles what types of issues. The tablets have frequently lost connectivity or gone out of service, caus- ing missed trips and anxiety. AC Transit notes that the vendor has been good about respond- ing and resolving issues immediately, but the agency believes these issues should have been explicitly determined in the contract. Types of Vehicles Used for General Public DRT/Microtransit Service With few exceptions, transit agencies that responded to the survey utilize minibuses with capacities ranging from 12 to 26 passengers (Figure 2). The reasons behind using these vehicles are straightforward: • The capacity of these vehicles is more suited to the low level of ridership typically experienced with general public DRT (between two to 14 passengers per hour). • The smaller size of the vehicles makes them easier to maneuver than large buses, allowing them to access areas that larger buses cannot in neighborhoods with more narrow streets. • The smaller size of the vehicles makes them more acceptable in residential areas of lower density that would most likely object to large buses operating on their streets carrying few people. • The minibuses are more fuel efficient and less costly to operate. • The minibuses are often the same or similar to the vehicles used by transit agencies or their contractors to provide paratransit services for people with disabilities. This provides agencies with a bit more flexibility when DRT vehicles need service or break down and sometimes reduces the need for separate inventories of parts. This can also result in a lower contract rate for those agencies using private operators that provide both paratransit and general public DRT. Source: North County Transit District. Figure 2. Example of the cutaway minibus used for general public DRT by many agencies.

30 Microtransit or General Public Demand–Response Transit Services: State of the Practice • A number of transit agencies reported that operators of minibuses under a certain passenger capacity do not need a commercial driver license, making it easier for the provider to attract and retain vehicle operators. • The minibuses are large enough to carry multiple passengers who board at transfer stations, schools, or employment centers. Table 3 summarizes the characteristics of the vehicles used by the respondents. A number of agencies noted that they did not need APCs since they have the boarding and departure data available from their automated scheduling information. There are a few agencies that reported using vehicles other than the fairly standard type of cutaway minibuses noted in Table 3 or that have agreements with other services if capacity is an issue during peak times. HART’s contractors started by using small Ford Transits [KCATA utilized larger versions (Figure 3)] and Chrysler Caravans. In May 2017, HART added four Tesla X Model (sedan) vehicles into the service plan. Innovation District in Tampa Bay donated the Tesla vehicles to HART. The fact these vehicles came with no cost made this an irresistible offer, and their all-electric nature helped to add a bit more buzz to marketing the new DRT services. Although the Tesla vehicles were not accessible, the other vehicles that were accessible could be dispatched to any of their four DRT zones when needed. NVTA’s contractor uses cutaway minibuses in three of their Flex zones and a trolley in the fourth Flex zone (Yountville). However, NVTA has ordered 30-foot low floor electric BYD buses for use in three of their four Flex zones for the future. Agency Manufacturer Length (feet) Passenger Capacity Wheelchair Capacity Low Floor AVL APCs WiFi AC Transit El Dorado Aerotech 24 16 2 No Yes Yes No Cherriots Starcraft 27 14 4 No Yes No No DART (Dallas) ARBOC 26 14 2 Yes Yes No No Denver RTD El Dorado 23 14 2 No Yes No No Greater Dayton RTA El Dorado Ford E - 450 22 8 4 No Yes No No HART Ford Transit Chrysler Caravan 14.5 4 1 Yes Yes No No Houston METRO ARBOC 24 12 2 Yes Yes Yes Yes KCATA Ford Transit 22 16 0 No Yes No Yes Kitsap Transit StarTrans 24 14 2 - 5 No Yes No No LYNX ARBOC Chevrolet 23 14 2 Yes Yes No No G4500 MST Starcraft 24 17 2 No Yes No No NVTA BYD 30 26 2 Yes Yes Yes Yes NCTD Supreme 20 15 2 Some Yes Yes No SacRT* Starcraft 27 22 2 No Yes Yes Some SJRTD Glaval 26 19 3 Yes Yes No No TDU Glaval 27 14 1 No Yes Yes No Note: ARBOC is a manufacturer of low floor buses of various sizes for the public transit industry. *SacRT uses Starcraft vehicles whose fuel is compressed natural gas. Table 3. Features of the vehicles used by respondents for providing general public DRT service.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 31 The Greater Dayton RTA utilizes their paratransit vehicles to take advantage of existing resources to minimize capital or operating costs. RTA did not need to add additional vehicles for their DRT service since customers comingle with their paratransit services. ADA trips are given priority, but anyone can ride on these vehicles as long as space is available and riding does not hinder the service requirements or quality of the service for ADA-eligible customers. When space does not exist, customers are offered a choice of using Lyft, Anton’s Taxi, or Secure Trans- portation, all of which are offered free of charge to passengers. The Stark Area RTA intends to utilize its existing fleet of paratransit vehicles, which include cutaway buses and MV-1. Similar to the Greater Dayton RTA, the agency anticipates all of their demand–response vehicles will be utilized throughout the day for general public DRT when not completing an ADA trip. Cutaway buses, often referred to as minibuses, are vehicles built with a bus body on top of a truck chassis such as those made by Ford or Chevrolet. They typically have a wheelchair lift and seat up to 20 passengers. Training Necessary for Transit Agency Staff The level of effort associated with training in-house staff in agencies that provide general public DRT varies considerably from one transit agency to another. As noted earlier in the report, private contractors that have experience in paratransit service delivery and, recently, in general public DRT/microtransit provide the majority of DRT services. In such cases, the in-house role can often be simple contract oversight. For instance, HART made quality assur- ance calls and on-board checks through secret shoppers to monitor the contractor’s perfor- mance and to help ensure that service delivery met HART’s standards. They also had to train administrative staff to reconcile invoices from the contractor and check to ensure that every- thing associated with the contractor’s performance is correctly documented. MST, NCTD, and SJRTD also reported that their private contractors have experience in the provision of both paratransit services and DRT and are responsible for their provision, while transit agency staff have oversight for contracting. The contractor is responsible for training their staff who are providing the service in their communities. Other agencies noted that even though a private contractor might operate the buses on the street, the transit agency staff is responsible for scheduling pickups and communicating with Figure 3. Ford Transit model used by KCATA for microtransit services.

32 Microtransit or General Public Demand–Response Transit Services: State of the Practice customers. At LYNX, agency training consists of having new employees sit with an experienced employee to learn how the software for their scheduling and ride reservations application (app) works. They claim the DoubleMap software app is intuitive and does not require a great deal of training. Dispatch staff are required to ride Flex routes to understand the vehicle operation, become familiar with the service area, and learn landmarks. RTD has been developing, implementing, and managing Call-n-Ride services for almost 15 years. With 21 Call-n-Ride zones and more than 40 general public DRT minibuses to oversee, several staff are assigned and have developed procedures and skills for this service, including performance evaluation and reporting. The experienced staff members train new employees. DART (Dallas) also has many years of experience (since 2006) in contracting for On Call demand–responsive service. While the challenge of managing DRT is not new, the agency is moving forward with providing service in eight new zones using new technology. For the new service called “GoLink,” planning, finance, mobility, and marketing staff are all involved in preparations and participate in the software and operational testing prior to making the service operational. The staff received training from the app developer (DoubleMap) and conducted dry run testing in a live format prior to initiating service for the public. This preparation took over a month and resulted in major changes in policy and some changes in the software, based on the dry run testing. NVTA did not go through extensive training when switching over to on-demand services. Before the services became solely on-demand, they were operated as deviated fixed route ser- vices. Dispatch staff was generally familiar with taking trip itineraries and dispatching those trips to the drivers if they needed to make a deviation. This incremental change allowed for a seamless change over. The Transit District of Utah understood that scheduling demand–responsive fixed route deviation service presents different challenges from scheduling traditional fixed route service and chose to use paratransit schedulers to schedule the demand–response route deviation services. TDU developed specific training, which included schedulers understanding both the fixed route aspect of their route deviation services and the methods for accommodating requests for deviations. Houston METRO initially contracted for their DRT dial-a-ride service but decided to bring the service in house. To facilitate a smooth transition and to decrease the need for training, staff from the contractor were hired as METRO employees. METRO also used existing Trapeze PASS 16 software, which was familiar to their paratransit scheduling staff. The Greater Dayton RTA reported that the transition to providing DRT services was made less difficult by their extensive experience with providing paratransit services that also require advance reservations and scheduling that are handled through Trapeze PASS soft- ware. Training in scheduling DRT services was provided to all quality service representatives in advance of the start of service by creating mock customers and booking trips accordingly. Service delivery is either by agency personnel utilizing paratransit vehicles or by a partnered provider. In-house staff entirely operates the Flex service of AC Transit. The only contracted piece is the scheduling and ride-matching software. Staff began training operators 3 months in advance of the new service launch. DemandTrans, a technology company providing scheduling and res- ervations software, had their staff on site for the initial training and launch. AC Transit staff wrote an operator training manual based on information supplied by DemandTrans, as well as internal procedures developed by staff in consultation with operations and training staff. The initial training combined 8 hours of classroom training with 8 hours of drive time per operator.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 33 The classroom training consisted of reviewing the operator training manual, procedures, and mobile tablet operation. The driver training consisted of running dummy trips from stop to stop. Subsequent training for new operators has consisted of 4 hours of classroom training and 4 hours of drive time. AC Transit believes that operators tend to forget what has been covered in the classroom and benefit most from operating the service while being shadowed by an expe- rienced operator. At Cherriots, a pilot project overview was provided to all staff in the agency to enhance their general familiarity with the new service paradigm. The departments most directly affected by the new service were Operations and Customer Service. Both departments were directly involved in the planning and implementation of the project. Planning department staff led the training. The agency reported that all departments became familiar with the agency’s expectations by the first day of the service. Marketing and Communications was also involved with the pilot program to enable the department to conduct an extensive public education campaign to ensure that customers were ready to use the service. Sacramento Regional Transit reported that training and preparation were minimal because similar service was already in place, and the only significant change was the use of new schedul- ing and reservations software to allow for increased dynamic service. The agency reported Trans- Loc trained the operators and dispatchers in how to use the software in a few hours and claims there was no IT–related preparation required because the system is cloud based. Technology Used to Make General Public DRT Efficient and Customer Friendly One of the survey respondents (DART in Des Moines) reported that they still schedule their general public DRT trips with paper and pen. Another survey respondent reported that the small city of Escalon outside Stockton, California, provides dial-a-ride DRT from 8:00 a.m. to 5:30 p.m., with no special software, but carries only approximately 20 passengers a day. In the 1980s, cell phones started to become popular and provided people with the opportunity to call for transportation services from wherever they were. Operators in vehicles could receive calls as well. In the late 1990s, there were transit systems providing fixed route deviation services. By 2004, a few transit agencies were establishing Call-n-Ride services that were the first practical forms of more dynamic general public DRT. Passengers could call the vehicle operator directly and request a ride, which the driver would process by writing the pickup and drop-off points on a clipboard while using hard-copy maps to find the locations. This old-fashioned system worked at first with a low level of demand, but it was clearly not scalable to handle multiple and nearly simultaneous requests for service. It also became unwise and increasingly unsafe to use a cell phone while driving a vehicle. By 2005, a multitude of technology advances were made that led to smartphones, a mobile internet, GPS, Google maps, interactive voice response, Voice over Internet Protocol, and software as a service. Suddenly there was an ability to assemble platforms and applications and to use cloud-computing resources. All these advances and more made it possible for transit agencies to • Improve customer service through automated booking and provision of estimated times of arrival. • Automate scheduling to enable dynamic service configurations. • Promote productivity and lower costs. • Automate the manifest for mobile service delivery, data capture, streamlined back office processes, real-time supervision, and facilitated service planning.

34 Microtransit or General Public Demand–Response Transit Services: State of the Practice Transit agencies interested in more dynamic scheduling and reservations can now engage in general public DRT/microtransit more easily than ever before. A number of technology compa- nies have sprung up in the past few years offering technical assistance to transit agencies inter- ested in pursuing real-time, customer-centric DRT. Among the newer companies mentioned by survey respondents are DemandTrans Solutions, TransLoc, DoubleMap, VIA, RidePilot, Luum, and Spare Labs. RTD is the transit agency that has served as the industry pioneer in pursuing the advantages offered by these new technological advances. The description for the technology platform used at RTD is from a paper published by the agency’s director of service development and their technology provider and is excerpted below: The technology platform must represent a comprehensive “operating system” for all forms of DRT oper- ated by the agency. The most fundamental purposes of this DRT operating system are the automation of the order taking and vehicle scheduling functions, which must be compatible with the Call-n-Ride program’s service delivery model that is driver-centric, customer-focused, includes a spectrum of DRT modes of oper- ation, and avoids call center functions such as agent-based reservations or dedicated human dispatchers or schedulers. Specifically, this meant that the following functionality needed to be incorporated into the technology platform: • A mobile device-based driver application that enables driver-based order taking, automated and semi- manual scheduling, trip manifest management, map-based trip routing assistance, and automated data communication • A Web-connected, GPS-enabled, mobile device (ruggedized tablet computer or smartphone) capable of 4G or higher speed wireless data communication, to host the mobile application • A Web-based customer self-service reservations application; the combination of self-service reserva- tions and driver-based order taking should be capable of handling all normal trip requests • For maximum flexibility, an agent-based version of the customer-based Web reservations application, available for such purposes as multi-agency coordinated services • A fully automated scheduling application, with appropriate algorithmic capabilities, to handle the scheduling requirements of any DRT service configuration used by Call-n-Ride—many to many, feeder, checkpoint, flex-route, etc. • An automated customer notification system that supports all relevant communication channels, namely text messaging, e-mail, and automated voice, to inform customers of upcoming trips and to provide advance notice of when a vehicle is due to arrive—or to notify the customer that it is running behind schedule • Interactive voice response capabilities (including limited speech recognition) for both trip ordering (reservations) and advance notice of previously booked trips, with ability to cancel the trip • Mechanisms to track the vehicles in real-time (displayed on a map) and to enable the service provider(s) and the agency managers to continuously and immediately view the status of operations in each service area; i.e., support real-time operational supervision • Automated data collection—using both the mobile technologies and centralized server as data collection points—that feeds a data warehouse that supports customizable analysis and reporting • Mechanisms that enable key system parameters to be automatically or semi-automatically updated based upon the operational and transactional data collected and processed by the system • Ability to scale from a single agency and DRT service to a metropolitan-wide network of participating services and organizations which can remain independent; the Denver RTD wished for this technology platform to support future regional efforts for coordinated services. Notably absent from this list is any requirement for human-assisted centralized dispatching, which is a standard feature of paratransit software packages. Given the driver-centric nature of Call-n-Ride and the small number of vehicles in each service area, there did not appear to be a compelling need for such functionality (p. 63, 5). Ridesharing company VIA issued an RFP in 2018 for a pilot Mobility on Demand Rideshare program. The language describing the technology platform it will require is included in Appen- dix B, which contains the entire RFP. DART’s original On Call service utilizes software from Trapeze, a company that provides soft- ware solutions and services to the public transportation industry. The new GoLink service will

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 35 use an app-based software provided by DoubleMap that requires a smartphone. Customers without smartphones can call a telephone number, and the person answering the phone will enter the trip request into a webform. The average time it takes for an on-demand vehicle to respond to a request for a trip is 11 minutes using the new app. The agency’s intent is to use the DoubleMap app for both the On Call and GoLink services in the future. When consider- ing what technology company to work with, DART determined that DoubleMap had software that was operational and would not require significant modifications to meet their deadline for the new service. NVTA also uses an app provided by DoubleMap that is considered a white label version of their ride-hailing app named “TapRide.” This app, dubbed “Ride the Vine” was implemented in December 2017. Since it began, almost 30% of rides are scheduled on the app, which provides another outlet for passengers to request rides other than calling dispatch to schedule service. The backend of the app allows dispatch staff to quickly schedule rides over the phone and send the details to the driver. Many of the responding agencies (e.g., Greater Dayton RTA, Houston METRO, MST, NCTD, SJRTD, and TDU) utilize software and technical assistance from Trapeze, a firm well known for ADA paratransit software. LYNX on-demand services were initially provided using Trapeze PASS (paratransit schedul- ing software), but once trip volume and routes increased the agency transitioned to Trapeze FLEX (Flex route scheduling software). In 2018, the agency had a smartphone application devel- oped by DoubleMap that allows customers to request, check on, or cancel a Flex service trip. Customers without smartphones may call a dispatcher to request a trip. MST’s demand–response service and vehicles are tracked using Trapeze software. Customers call into a toll-free number to schedule a trip. While the agency does not have an app or technol- ogy to allow demand–response customers to make their reservations through automated means, the feedback has been positive regarding their current system. NCTD uses Trapeze software for scheduling and dispatching its Flex and ADA paratransit services and believes it is accurate and efficient. Customers do not need smartphones to make reservations. If the customer needs a route deviation to reach the final destination, the customer is encouraged to call a reservation line from any phone at least 30 minutes in advance of the trip. TDU uses Trapeze for both paratransit and their route deviation services, first using Trapeze FX and then copying the data to Trapeze PASS for the deviation portion of the service. The agency also partnered with RidePilot and Cambridge Systematics to enhance and develop open source software that allows smaller agencies to use this software to schedule exclusive demand–response service. SJRTD uses app technology provided by Uber and TransLoc for their RTD GO! route devia- tion service. Those without smartphones or credit cards can call RTD, and the agency will schedule the ride for them. The Sacramento RTD worked with TransLoc to establish a small pilot project within the agency’s timeframe. The technical assistance received cost $25,000, for which the agency received licenses for up to 10 vehicles, training, and cloud tech system access. The agency claims no IT preparation was required to accommodate TransLoc’s technology. The service started as a pilot in February 2018, witnessed a 50% increase in ridership over previous dial-a-ride services, and expanded to serve two additional communities within 2 months.

36 Microtransit or General Public Demand–Response Transit Services: State of the Practice GCT is also using TransLoc for its pilot program that began in September 2018. GCT is a small agency with only five full-time employees and as agency director Karen Winger stated in her response to this TCRP survey, “Any place we can use technology to make things easier, I want in!” The agency will continue this pilot until April 2019 and then go out for a competitive procurement for services after the pilot. Customers will continue to be able to make reservations via smartphones or regular phones once the project transitions from pilot status to permanent status. To date, almost 90% of reservations are made via the app. Gainesville RTS in Florida also intends to use TransLoc in the future for a pilot serving low-income areas with disconnected street patterns. As noted earlier in this report, it is important to realize that general public DRT/microtransit is not intended to be a taxi or a TNC service. While microtransit might have software with simi- lar capabilities to make same-day reservations, microtransit does not have the resources in the form of numerous vehicles in one zone to serve great numbers of individual trips as taxis or TNC services do. General public DRT/microtransit uses only one or two vehicles in a service zone to respond to a multitude of trip requests. Consequently, the time it takes for the general public microtransit vehicle to arrive at a requestor’s location is dependent not just on the capability of the scheduling software but also on the number of vehicles in the zone that can be dispatched, the size of the zone, and the volume of requests that are made. The algorithms of the software will optimize the routing of the vehicle to serve all requests possible while still meeting time points in the vehicle’s schedule. The important factor is not how long it takes to respond to a single request for service (it could be 5 minutes, it could be 30 minutes or more depending on circumstances at the time) but how closely the vehicle arrives at the time it indicated to the passenger that he or she would be picked up. Basic Performance Metrics and the Business Case for General Public DRT/Microtransit Service Transit agencies fully understand that general public DRT will not perform at the same effi- ciency levels as fixed route transit. The demand for transit service is not large enough in areas of lower density to result in a high number of passengers per hour, even with the greater flex- ibility offered by DRT. However, there is still a business case to make to support the provision of general public DRT as follows: • The cost per passenger is considerably higher than that of fixed route service due to the rela- tively small number of passengers carried, but the cost per hour of service provided is usually considerably lower due to the use of contractors and smaller vehicles. • The cost per passenger of general public DRT is considerably lower than the cost per passen- ger for traditional paratransit services that have inherent inefficiencies; hence, the relatively small investment in recent technology to support more dynamic DRT is worth the cost. • The price of technology is decreasing rapidly. LYNX reported that their procurement of the technology necessary to provide dynamic demand–response service initially took 3 years and more than $800,000 when they started the process in 2014. SacRT reported they paid only $25,000 to TransLoc to utilize the already-developed software they needed to activate their microtransit service in 2018. Table 4 shows the ridership performance of the DRT services provided by transit agencies that responded to the survey. Respondents reported an almost remarkable consistency in average ridership ranging from 2.4 to 4.7 passengers per hour. As is the case with fixed route services, DRT that operates in areas with greater population or employment density tends to perform better than DRT operating in areas of lower density.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 37 In Call-n-Ride or Flex zones with employment centers or rail stations, agencies reported DRT carrying as many as 9 passengers per hour on average throughout the day, with higher num- bers during peak hours. DRT that functions primarily as feeder services rather than circulator services generates higher ridership per hour due to a greater number of spontaneous boarding at stops that generate or attract larger numbers of riders such as rail stations, transit centers, schools, or employment centers. For example, among the Denver RTD’s DRT services that are predominantly many-to-many operations (with either no cycle points or 60-minute frequency for cycle points), service productivity is typically in the range of 3 to 4.5 passenger trips per vehicle service hour. In contrast, in service areas that use structured DRT (and which typically have cycle times of 30 minutes or less), productivity typically ranges from 5 to 9 passenger trips per hour, or approximately 50% more (5). The average cost to provide an ADA paratransit trip in 2010 was $29.30, as reported by the U.S. Government Accountability Office, as compared with the average cost of $8.15 for a fixed route trip. The average cost to provide an ADA paratransit trip increased 10% from 2007 to 2010. See ADA Paratransit Services Demand Has Increased, but Little Is Known About Compli- ance: Report to the Committee on Banking, Housing, and Urban Affairs, U.S. Senate, Nov. 2012. https://www.gao.gov/assets/660/650079.pdf. There is great variability in the cost per passenger associated with general public DRT. TDU reported the lowest cost per passenger at $7.34 for a service that is route deviation in nature, while AC Transit reported the highest cost per passenger at $71.00. The average for all systems reporting was $23.56. If the lowest and highest cost figures were eliminated—as well as the cost figures for HART, whose service was discontinued due to the contractor’s inability to continue Transit Agency Contract or In house Cost per Vehicle Service Hour Passengers per Vehicle Service Hour Trip Cost per Passenger AC Transit In house $214.00 (fully allocated) 3 $71.00 Cherriots In house $65.00 3.5 $18.57 DART (Dallas) Contracted. DART provides vehicles and facilities but not fuel. $46.00 2.5 for original DRT service and 3.5 for new GoLink service. $18.40 $13.14 Greater Dayton RTA In house and contracted RTA pays Lyft and taxis and uses in-house paratransit. Not applicable $13.00 Denver RTD Contracted $83.00 3.8 $21.84 HART Contracted HART pays contractor by trip and not by hour. 3.5 $10.00 Houston METRO In house $75.00 2.4 $31.25 Kitsap Transit In house $130.72 3.66 $35.68 LYNX Contracted $41.17 3.3 $12.60 MST Contracted $54.18 4.03 $13.44 NVTA Contracted $44.48 2.6 $17.00 NCTD Contracted $97.00 2.7 $36.00 TDU Contracted and in house $34.69 4.7 $7.34 Note. The numbers are self-reported figures from agencies that responded. Table 4. Ridership per vehicle service hour and costs per hour and per trip.

38 Microtransit or General Public Demand–Response Transit Services: State of the Practice providing service at such low rates—the average cost per passenger was $21.70. This is approxi- mately 2.5 times the cost per passenger for fixed route service but usually more than 30% less than the cost of traditional paratransit (26). The Greater Dayton RTA was able to confirm its business case for DRT when it demonstrated to the board of directors that the cost per passenger had been $133.00 per trip when serving the areas with fixed route service compared with the costs of $13.00 per trip when serving the same areas with DRT. The farebox recovery ratio associated with DRT services also varies. In some cases, the service is free and results in no direct farebox return. DART in Dallas reported a farebox recovery ratio of almost 12%. RTD and NCTD reported approximately 7%, while SJRTD realized a 5% fare- box recovery ratio. Respondents reported other miscellaneous sources of revenue for their DRT services. NVTA and LYNX receive operating support from the cities in which the DRT services are provided. A number of agencies reported receiving operating support from a variety of state and federal grants. LYNX receives FTA Section 5310 and Section 5317 New Freedom grant support, while San Joaquin receives support from the FTA Section 5311 (Formula Grants for Rural Areas) program. HART and LYNX received service development grants for the first years of their pilot projects from the Florida Department of Transportation. HART achieved one of the lowest costs per trip based on the provisions of their contract that called for the contractor to be paid $10.00 per trip instead of a certain amount per service hour. If the trip started and ended within the Flex zone and did not involve connecting to the regional transit system, the passengers paid $3.00 to the contractor and HART paid only the remaining $7.00. Trips that connected to HART at designated stops were priced at $1.00 to the customer, with HART making up the difference up to the $10.00 rate. Unfortunately, the contractor for HART advised that it would require raising the rate of payment per trip from $10.00 to $24.00 in order to sign a new contract. HART’s CEO did not believe he could make a good business case to continue the service, and the board of directors agreed to terminate the program on July 31, 2018, and provide modified fixed route service in the Flex zones. DART in Dallas indicated their ideal operating model would be some basic level of contracted service supplemented by purely independent contractor shared-ride TNC service, in which the agency pays by the mile or the mile plus time. It frustrates DART to pay by the hour even if no one requests a ride. In the long term, the agency hopes to use autonomous shared-ride services to reduce the labor costs and offer more service at lower costs to their customers. RTD is con- sidering the use of non-dedicated vehicles (TNC services or taxis) for off-peak hours in certain zones that are commuter-oriented and generate few requests for DRT service in the middle of the day. This could bring the average cost to serve such Flex zones down to close to $12.00 per trip in such zones. Many of the survey respondents noted that they enforce a no-show policy that allows the agency to preclude a passenger from using the service for a set amount of time if the passenger failed to be present for a trip that was reserved. For instance, if a person fails to show for four requested rides, he or she can be denied service for the next month. Transit agencies recognize that DRT services perform at lower efficiency levels than fixed route and do not directly compare these two different services with each other when determin- ing what routes should be retained, changed, or eliminated. Flex routes tend to fill the coverage part of the traditional ridership versus coverage service dichotomy, as well as offer mobility and transportation equity for riders who may otherwise be deprived of service if strict ridership standards are applied. Nevertheless, DRT service is still public transit and most agencies believe DRT service should be evaluated by many of the same metrics used to evaluate fixed route ser- vice, including ridership and cost per rider. TDU measures their Flex routes in relation to each other on an ad-hoc basis, and some routes with the lowest ridership and highest rider subsidies

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 39 have been eliminated or have had trips cut, when they could not be justified by other measures or when the agency determines the resources can be better used elsewhere. Lessons Learned All survey respondents were asked the following questions: “What have you learned that would cause you to have done something differently (or will do differently) in providing general public demand–response service? What are the most valuable lessons learned that you would want to share with other agencies that are considering establishing similar services?” A few answers from different agencies contradict one another but this is an expectation given the highly customized circum- stances general public DRT services address. The answers provided are categorized as follows and should be useful for any transit agency that is considering offering general public DRT/microtransit. Goal Setting and Ridership Expectations • Set realistic goals. This is a low ridership service for low-density and low-demand areas or times. • Establish clear and measurable objectives. • Allow seven passengers per service hour, which approaches the limit of service productivity. Trip denials start to happen above this. • Ensure that your staff and board are educated in the level of service you are attempting to achieve and why the current business model needs to change. Operations Considerations • Set small service zones. When the structure of a service zone is around a schedule point at a major hub, the service zones should be in the range of 5 to 7 square miles. • Plan for the market. Assess potential ridership, customer characteristics, and service area char- acteristics. Research customers’ travel patterns and configure the service to meet their needs. • It is hard to do both on-demand services and subscription services and reach a proper balance. The trips booked ahead limit the ability of the system to respond to on-demand requests. However, it is worth pursuing both because limiting one of these options limits productivity. • Ensure that time connections are met. • App-based scheduling is more effective than general demand–response scheduling software. Dynamic technology is necessary for efficient operation of Flex services. Phasing In Service • When introducing new service, it might be practical to allow old service to run, but 3 months of simultaneous service operations are sufficient. On the other hand, it might be best to dis- continue redundant fixed route bus service first. The external and internal politics of discon- tinuing fixed route bus service have been significant, and there is little likelihood of addressing the politics in the future. In this sense, it is easier to establish pilot projects in which there is no transit service. If they are successful, it then might be easier to make the case to replace poorly performing fixed route service with DRT service. • It is wise to offer a pilot demonstration service so that skeptics and critics have the opportunity to examine the concept in the field before extensive commitments are made. • Replacing a service is easier than eliminating a service. Opposition to eliminating the fixed route service (which given its productivity would have been justified) would have been greater without offering Flex service as mitigation and replacement. It was beneficial to launch the service in an area with existing transit service to draw from an existing customer base.

40 Microtransit or General Public Demand–Response Transit Services: State of the Practice Marketing and Customer Service Considerations • Marketing promotions are even more difficult for DRT than for fixed route services. Extensive customer outreach is imperative in making service successful as well as substantial coordina- tion with the communities affected. The greatest challenge is teaching people how to use the service. It is often difficult for current customers and potential customers to translate the description of the concept into a positive experience. • Offer free travel training for paratransit customers transitioning to flex services. • Understand the importance of the more personal customer service aspect of on-demand service. Using drivers who have driven paratransit before helped, because they were more sensitive to people’s needs for demand–response service. • Use the same fare structure and fare media as fixed route services. • Have more booking options for customers (standard phones, smartphones, walk-on boarding, and subscriptions), which results in more productivity. Contractor Considerations • For the future, the transit agency will ask the vendor to provide the vehicles and will depend more upon shared ride TNCs for service rather than a single provider. The only worrisome issue with the TNCs is providing adequate wheelchair capacity. • In the future, our agency would like to pay by the mile of service provided. Shared ride TNC service would allow our agency to pay only for the service that transports people and not pay by the hour, as occurs under contracted operations when the transit agency pays even if no one asks for service. If our agency can purchase from the private market in a way that we pay only for the service we need to provide a trip, we think we will be able to offer more service with lower costs. However, an agency must be able to address ADA requirements. • The transit agency should have control over the mobile application data. When the contractor owns the data, this ownership does not allow for proper monitoring of service. New contracts will request a dashboard for monitoring of data. Barriers to the Provision of General Public DRT Services The survey contained the following question: “What were/are the barriers to providing gen- eral public demand–response service with your own vehicles and personnel or through tradi- tional private transit services providers?” The responses of survey respondents reflected different experiences based on their particular circumstances, and each identified barrier might not be present in other locations. The answers provided are categorized as follows and should be useful for any transit agency that is considering offering general public DRT. Human Resources/Personnel Issues • The collective bargaining agreement can inhibit flexibility and nimbleness. • Strict interpretations of the collective bargaining agreement might require all 1,300 drivers to train for a job only six drivers will do as DRT operators. • Many senior drivers are not tech savvy and do not interact well with the technology. • It would be difficult to serve existing routes in addition to a new microtransit or demand– responsive service due to absenteeism, high turnover, and a slow process to hire new bus operators. • The biggest barrier to providing service through private contractors is managing their per- sonnel and reminding the contractors they are ambassadors and representatives of the public transit agency.

Survey Results from Public Transit Systems That Have Implemented General Public Demand–Response/Microtransit Service 41 Technology Issues • Having the requisite knowledge as it pertains to technology assessment and acquisition. • Equipment is prone to failure and there must be sufficient training from the vendor. • Integrating data from the scheduling software with APC, AVL, or other administrative systems, which can be a challenge. Costs and Financial Resources • The primary barrier was the capacity to accommodate the demand in the Flex zone with one vehicle. Adding additional DRT vehicles raised the cost back to where operating fixed route became the more cost-effective model. • Providing the vehicles through the public agency has tended to increase the capital cost of the vehicles, due to adding all the technology and compressed natural gas fuel requirements. This increased cost would not have been the case if private providers had provided the vehicles unless, of course, required by the transit agency. • Funding and the effective coordination of all available providers as we “blur the lines” of mobility. • The main barrier in providing general public demand–response service is the ability to com- pete with TNC services like Uber and Lyft. Customers using a TNC are able to open an appli- cation and have a vehicle at their door within minutes that can take them anywhere from a half a mile to 30 miles and beyond. Our agency does not have the financial resources or the fleet size to provide on-demand services at the level of TNC services. Miscellaneous Factors • Another factor is the location of operating facilities in relation to the location of the DRT zones and the amount of space transit agencies have at the operating base to expand the fleet. • The transit agency hopes to modify the DRT program to open it up to TNC services. There is a federal restriction for using TNC services as part of our DRT services because of ADA issues. There is a spot bill in the legislature that would require TNC services in California to be equipped to accommodate wheelchairs. This requirement will create additional flexibility for how the agency deploys on-demand services in various parts of the service area. • Developing performance measurements and assessments for general public DRT service. • Customers, staff, and stakeholders do not know much about DRT; they just want bus service. Customer and Community Feedback Survey respondents were asked to answer the following question: “Please describe the type of feedback you have received from the public regarding your general public demand–response services,” to find out how their DRT services have been received by customers and the com- munities they serve. In general, the feedback transit agencies have received for their general public DRT/microtransit service is positive and, in many cases, quite enthusiastic. There have been some complaints where fares are higher for DRT than for fixed route services and about more attention given to first mile/last mile trips than to requests for trips to other destinations within the zone. The responses should not cause transit agencies to fear there is a negative perception of DRT services by those who have used them. Because each area provides service a little differently and some areas have more detailed data to support their findings than others do, each agency’s response follows: • AC Transit. Feedback from customers has been mostly positive. During the pilot year, staff received just three complaints regarding the elimination of the fixed route. Ninety-four percent (94%) of riders surveyed preferred Flex over restoring the fixed route, and 70% of riders said

42 Microtransit or General Public Demand–Response Transit Services: State of the Practice they would take AC Transit more if the service were expanded. The service has generated sig- nificant attention and interest from the riding public, media, elected officials, and peer agencies. • Cherriots. Online and phone surveys conducted by the agency indicated that customers’ overall experience with The Connector DRT service was rated very good by 55%, good by 32%, poor by 4%, very poor by 0%, and neutral by 9%. • DART (Dallas). The original On Call service has received positive feedback, but with some complaints that the service does not offer enough coverage and vehicle availability. The service is highly prioritized for first mile/last mile trips. Some customers have been unhappy about lack of availability to travel to destinations other than to the rail station during some service cycles. • DART (Des Moines). Little feedback. There is some concern on levels of service. • Denver RTD. The high levels of automation within the system make relatively hassle-free on-demand service feasible for customers, who give Call-n-Ride a rating of 4.6 out of 5, which represents RTD’s highest satisfaction rating among all of its modes. • Greater Dayton RTA. Positive feedback from customers using RTA Connect On-Demand. They are receiving increased service and accessibility to main line services as a result of offer- ing this service 24 hours per day. • HART. Positive feedback from our users. There were some complaints on wait times during rush hours when heavy traffic delays response time. • Houston METRO. The feedback from the customers who use the service has been overwhelm- ingly positive. People are thrilled with the options of either calling or catching the bus at time points. Customers regard the DRT service as one of the better service offerings they have. The feedback from those citizens who have not used the service ranged from skeptical to strongly opposed initially, but there is evidence of some interest in communities that had originally rejected the service concept outright. • LYNX. Initially, when low-performing fixed routes were replaced by Flex services, custom- ers were hesitant to use the service. Over time, as customers learned to navigate the service, customers have come to appreciate the convenience of the option. • MST. Overall feedback from the public for the demand–response service has been positive. MST has also received positive feedback from businesses like medical facilities and has been asked to expand service to additional destinations. • NCTD. Many customers requested expanded Flex service coverage and span. Other customers expressed opposition to the October 2017 removal of FLEX 373 and 374 in Carlsbad, Encinitas, and Solana Beach. Prior to the launch of FLEX 392 and 395 (which replaced fixed route bus service), some customers were concerned about paying a pre- mium fare. • SacRT. The pilot has received a lot of positive feedback from both the customers and elected officials in the short time it has been in service. Two new communities are looking forward to receiving similar service. • SJRTD. There have been concerns from rural residents regarding the cost of RTD GO! and the lack of affordable general public Dial-a-Ride service in the rural areas. • TDU. In general, if customers do not understand the service model, they do not like the delay caused by the service deviation. Limited Saturday and Holiday service just began in 2017, which was funded via old Section 5310 funds, with positive feedback from our riders.