Use of Electronic Passenger Information Signage in Transit (2013)

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.

8 CHAPTER TWO LITERATURE REVIEW The literature review revealed many reports, papers, arti- cles, and press releases that have been written about the use of electronic passenger information signage in transit. This review has the following sections, including the five ele- ments identified in chapter one: • Underlying technology • Signage technology • Information characteristics • Information accessibility • Accuracy and reliability • Monitoring • Standards • Required resources • Decision processes • Selection criteria • Signage placement. The first step of the literature review was to conduct an online Transportation Research Information Services (TRIS) search. This TRIS search yielded more than 100 documents, the most relevant of which were reviewed and used as input to this report. The second step was to obtain and review articles, press releases, and website informa- tion directly from agencies and electronic sign vendors across the world. The third step was to review research reports from the FTA, FHWA, and TCRP. Finally, other papers and articles were obtained from different sources, including the following: • TRB annual meetings, • APTA conferences, • ITS America (ITSA) annual meetings, • ITS World Congress meetings, and • Internet searches. The References and Bibliography list all documentation reviewed for the Synthesis. UNDERLYING TECHNOLOGY The literature for this element of the research revealed that the technologies that are required to generate the information that is disseminated by means of electronic signage include “automatic vehicle location (AVL) software, computer-aided dispatch (CAD) software, software that calculates the real- time information from data generated by CAD/AVL systems and software” (1, p. 9) As stated in TCRP Synthesis 73 (4), an AVL system facilitates the “use of schedule adherence and/or location data to develop real-time predictions for bus arrival times at stops, and providing these predicted arrival times and other service announcements to the public using various methods.” Most of these underlying technologies have been the subject of numerous reports and articles, so this Synthesis will describe these technologies briefly in chapter three and provide references to these reports. Given the coverage of underlying technology in numer- ous research reports to date, the literature review focused on the most recent developments in generating information that is displayed on electronic signage. First, although AVL systems provide input to algorithms that predict when the next vehicle is going to arrive at a particular stop or station, one of the challenges associated with providing customers with accurate next vehicle predictions is the polling rate of the AVL system. “The polling rate of vehicles is often too infrequent to be of much use for real-time predictions. Many agencies are only able to poll their buses every two to three minutes, which can lead to inaccuracy in real-time arrival predictions. Often there is not enough bandwidth available for transit agencies to transmit vehicle locations at a higher frequency, even if the on-board system can transmit at higher rates” (5). Fortunately, many newer AVL systems use higher frequency polls (e.g., 30-second polls), since the cost of the necessary data communication (from vehicle to central dispatch) has been reduced over the past 3 to 4 years (2008–2012). Similarly, rail control systems that provide information to predict the arrival of rail vehicles must be reliable and accurate. “Data integrity of the core rail and bus operational systems that produce customer information, such as the rail control system, is important. Inaccurate data cannot support reliable customer information systems” (5, p. 62). In the United Kingdom, the Real Time Information Group (RTIG) reports annually on the status of real-time information systems deployment including the underlying technologies that generate the information disseminated through a wide variety of media (6). “At the end of 2011, 22,118 buses (50% of the total UK bus fleet of 44,057) were

9 fitted with on-bus tracking units and 3.3 [b]illion (64%) estimated bus passenger journeys occurred on equipped buses. The [real-time information] RTI equipped fleet now accounts for 50% of the UK bus fleet and 51% of the [Great Britain] GB bus fleet” (6, p. 11). Figure 1 shows the steady growth in the number of buses equipped with AVL in Eng- land since 2002. FIGURE 1 Number of buses equipped with AVL in England (6, p. 17). Several references discuss the use of underlying tech- nologies to provide electronic signage information about more than one mode of travel. For example, in 2010, elec- tronic signage at John F. Kennedy Airport in Jamaica, NY, was deployed to show passengers real-time information on AirTrain and connecting transit services [Long Island Rail Road (LIRR) and Metropolitan Transit Authority (MTA) New York City Transit subway and bus services]. “The signage displays information about approaching trains, the stops along a particular train’s route and what changes may affect the train’s service. Officials worked with the com- panies that operate the AirTrain and provide its SCADA (supervisory control and data acquisition) system to inte- grate digital signage into the train operations. ‘We now have certain triggers that work with the digital signage, so if we go into a bypass strategy the signage gets updated automati- cally. Now, for example, if it is a train that’s stuck and we run a divergent service, all of the information on the digital displays will get updated right away’” (7). Further, at Southampton Airport in Southampton, U.K., real-time multimodal traveler information is provided to arriving passengers on electronic signage. “The travel and transport information system obtains real time traffic and travel information from the Highways Agency, Southamp- ton City Council, Hampshire County Council, South West Trains/Association of Train Operating Companies and Red Funnel Ferries, as well as displaying timetable information for all local bus services and live [closed-circuit television] CCTV images for the local motorway network” (8). Figure 2 shows the system architecture. FIGURE 2 Southampton Airport traveler information display system architecture (8, p. 5).

10 Information is obtained from the data providers (shown on the left side of Figure 2) and transferred to the display system using several “standard communications protocols such as XML (eXtensible Markup Language) and Internet protocols such as SOAP (Simple Object Access Protocol). These data are then automatically formatted into a defined standard XML format by adapter software written in Java” (8, p. 3). To minimize capital and operational costs, one per- sonal computer is used to send the video to the LCD signs located throughout the airport, rather than using one com- puter per LCD display. In designing this system, the real- time nature of the information being presented in addition to the environment in which passengers are reviewing this information (e.g., while waiting for and claiming baggage, visiting restaurants and shops) led the system designers to rule out devices that would require passenger input (e.g., kiosks). “Therefore, by utilising large format LCD screens at key locations where passengers are carrying out these other tasks, they can be fed live information on train departures, bus movements and local traffic congestion” (8, p. 6). Less advanced underlying technology was used as part of a demonstration of a transport guidance system in a bus ter- minal in Tsukuba City near Tokyo, Japan. In this case, “The system combines a system which applies [radio frequency identification] RFID technology in wide use in distribution and an LED display type electronic display board. It detects a bus leaving the bus terminal and updates the display con- tent (the exit/entrance of this bus terminal is one location). On the premise that electric power will be supplied to road- side units, the on-board unit installed on the bus is only an RFID tag, while the road side unit is equipped with a reader/ writer” (9). Finally, underlying technology was described as part of the deployment of electronic signage for public transit in Johannesburg, South Africa, before the 2010 Fédéra- tion Internationale de Football Association (FIFA) World Cup. In this application of electronic signage for the bus arrival information in bus rapid transit (BRT) stations, an Advanced Public Transport Management System (APTMS) was implemented, consisting of in-vehicle and central fleet management systems. These underlying technologies pro- vide next-bus information (10). SIGNAGE TECHNOLOGY A considerable amount of literature covers display types (e.g., LED, LCD) and other characteristics such as what can be displayed using specific display types (e.g., characters only, characters and pictures). As summarized in (5), Intelligent Transportation Systems (ITS) summary deployment statistics from 2006 indicate that roughly thirty percent (30%) of transit agencies in 29 large metropolitan areas surveyed use [dynamic message signs] DMS in locations other than vehicles to disseminate transit routes, schedules, and fare information to customers. DMS, such as light-emitting diode (LED) and liquid crystal display (LCD) systems, show train destination, arrival, and departure information. When placed on loading platforms, they may flash to alert riders of an oncoming train or bus. In addition, [a]gencies across the United States use DMS and LED/ LCD monitors to communicate information to customers who are en route, on-board, or at-station. The use of DMS is more likely at heavy and light rail stations or bus depots than at bus stops; although dynamic signs are being introduced at major bus stops as real-time vehicle location information becomes more available. According to 2007 ITS deployment statistics, Alameda-Contra Costa Transit District in California utilizes DMS at 75 of its 6,000 bus stops. In contrast, the Bay Area Rapid Transit uses DMS at all 46 of its rail transit stations and is planning for the installation of in-station LCD screens to increase the type of information travelers receive. (5, p. 38) From Thessaloniki, Greece: The most common medium used for the distribution of real time [public transport] PT information, is the electronic display, also known as a Dynamic Message Sign (DMS). The dissemination of real time information is also possible through video monitors, interactive kiosks, personal digital assistants, telephones, Internet, and cable television. Passenger information can be made available on board, through Light Emitting Diodes (LED) and Thin Film Transistor (TFT) technology as well as on the wayside through various display technologies (LED single lines, LED matrix, LED lines, TFT screens). (11) (Note that TFT is a type of LCD display that improves image quality.) Historically, the early literature (1) covering electronic signage for public transit reported that most signs were LED, with LCD technology just beginning to be deployed. Currently, the literature reports a shift to full-screen dis- plays using LCD or plasma technology. For example, in the United Kingdom, the Real Time Information Group (RTIG) reported that “there were approximately 8,130 bus stops fit- ted with 3-line or multi line LED signs and a further 2,046 fitted with full screen (LCD or plasma displays)” (6, p. 22). As shown in Figure 3, the United Kingdom has a growing number of full-screen displays. This situation is discussed in more detail in chapter five, as this trend is making changes in the information that can be displayed on electronic signs. The display type of choice has remained the 3-line/ multi line LED type throughout the period 2002–2011. However, there is a long-term trend towards an increasing proportion of screens being full screen. Following a rise in 3-line/multi line LED displays between 2005 and 2008, growth in this display type has mostly remained flat between 2008 and 2011 around the 8,000 display mark.

11 Projections for 2012 and 2013 show a sharp increase in 3 line/multi line LED in 2012. Rises in LCD screens will be more modest. The proportion of signs which are 3 line/multi line LED has been at about 80% since 2010 and is expected to continue through 2013 (6, p. 22). FIGURE 3 Number of electronic signs in Great Britain (6, p. 34). Other systems abroad report the use of both LED and LCD signage. For example, in Milan, 42-inch [diagonal] LCD displays will be replacing the existing LCD displays (see Figure 4) in the subway (12). FIGURE 4 LCD display in Milan subway (Courtesy: Carol Schweiger 2008). In Rennes, France, STAR (Service des Transport en commun de l’Agglomération Rennaise) has recently introduced ‘INFOSTAR Synchro,’ a system that organises real-time information on screens on the platforms of the metro, at the main bus stops, and in the buses themselves. The information is also delivered via voice [to accommodate visually-impaired persons]. 70 screens have been installed on the metro platforms and at the entrances and exits of the ticketing halls in the 15 stations along the metro line. These screens let passengers know when the next two trains will be arriving. Work is underway to equip the busiest bus stops with LCD screens displaying timetable information in real time. The goal is to outfit 260 of Rennes Métropole’s more than 1,000 bus stops with these ‘passenger information posts.’ As with the metro, the screens installed at the bus stops inform voyagers of the arrival times of the next two buses. They can also be informed of any disturbances on the line. Some of which will even be solar-powered. (13) On-board buses, passengers are informed of the follow- ing (13): • The direction of the bus (terminus) • The four upcoming stops • Connections with other buses, the metro, or LE vélo STAR (bike-hire system) • Key public places on the line, such as city halls, hospi- tals, cultural and sports places, together with economic activity zones and commercial centers • The time remaining until arriving at the most impor- tant stops • Disturbances • Availability of LE vélo STAR on the line. Other types of electronic signage combine various types of display technology. In the Brussels subway, station signs display real-time information, including the route number and destination of the subway along with the number of min- utes until arrival and the location of other subways on that same route (see Figure 5). This sign has static (the route and station names on the map) and dynamic elements (dots dis- played under the current subway location and the real-time information displayed below the map). FIGURE 5 Real-time information sign in Brussels subway station (Courtesy: Carol Schweiger 2012).

12 advertising network (including passenger information) using LCD screens is being implemented in the 11 busiest rail stations in Belgium and 22 high-speed rail locations at Brussels-Midi and Antwerp-Central stations. Further, five 4 × 1 m horizontal screens are being introduced to provide advertising and information on weather, tourist opportuni- ties, station activities, and other information of interest to travelers (15). FIGURE 7 Touchscreen inside Brussels–Central Rail Station (Courtesy: Carol Schweiger 2012). Interactive screens similar to those installed in Brussels are being implemented in the Franklin D. Roosevelt metro station in Paris. “At the heart of the new [station] design, pas- sengers will discover two types of screens—16 in all—per- fectly integrated into the station. Three x 52-inch (132 cm) touchscreens with voyager information on each platform have replaced the typical paper-based supports. The screens recall a sort of giant iPad, with each one providing the fol- lowing content: The literature describes several sign systems in the United States and abroad that go well beyond displaying vehicle arrivals and departures—most of these new sign types are large touchscreen computers that allow user interaction. The New York MTA has deployed several 47-inch touchscreen devices called “On the Go! Travel Stations,” as shown in Figure 6. “On the Go! Screens were first unveiled in Sep- tember [2011] at the Bowling Green subway station in lower Manhattan and two major subway complexes: Atlantic Ave.- Pacific St. in Brooklyn and Jackson Heights-Roosevelt Ave. in Queens” (14). As of July 2012, the customer acceptance and reception to the signs has not been assessed. FIGURE 6 MTA On the Go! travel station (2). In Brussels, Belgium, eight large touchscreen moni- tors have been implemented in two major rail stations: Brussels-Midi and Brussels-Central (see Figures 7, 8, and 9). These 90 × 215 cm touchscreen devices allow users to obtain information about the rail station, real-time service disruptions, and timetables, as well as to plan a trip on the Belgian rail network. They are available in four languages (English, French, Dutch, and German) and are located in high foot-traffic areas. These devices and their use will be evaluated in the future to determine if any changes should be made to the content. To defray the cost of these dis- plays (€17,000 for each two-screen installation), a digital

13 • Maps of the metro, bus, and Paris surroundings • A map of the neighbourhood • A wayfinding tool, built to be intuitive and easy to use. It is a sort of modern version of the PILI (plan indi- cateur lumineux d’itinéraires), or light-based itinerary indicator map), the famous push-button map tool intro- duced to the Paris metro in 1937 • Five 82-inch (208 cm) communication screens per platform, managed by the RATP’s advertising arm Métrobus Publicité, for hosting ad campaigns and cul- tural content” (16). Another sign technology discussed in the literature is electronic paper display (EPD). “EPD technology is an elec- tronic sign capable of presenting text and images on a flex- ible surface that can be changed over time. EPD does not use a large amount of electricity. The technology has been in commercial use worldwide since 2005. It appears in elec- tronic books, cell phones, electronic billboards and other general signage. EPDs are touted for their superior read- ability and extremely low power consumption, compared to traditional LED DMS or LCDs” (17). Figure 10 shows an example of an EPD. In 2006, Hamburger Hochbahn AG, a rail company operating in the City of Hamburg, Germany, installed ‘mobile dynamic destination displays’ of traveler information using EPD technology as part of a pilot project to assess the applicability of the technology as an alternative to traditional electronic signage. There is limited information regarding the results of the Hamburg experiment. As of mid-2009, no U.S. transit agencies could be identified as currently using the technology. However, during interviews with information technology staff at Tri-County Metropolitan Transportation District of Oregon, it was mentioned that they [were] currently researching the technology for use in their system. (5, pp. 39–40) FIGURE 8 Touchscreen on outdoor platform at Brussels–Midi Rail Station (Courtesy: Carol Schweiger 2012). FIGURE 9 Disruptions displayed on touchscreen display inside Brussels–Central Rail Station (Courtesy: Carol Schweiger 2012).

14 FIGURE 10 Sample EPD (http://www.smh.com.au/news/ breaking/electronic-paper-that-bends/2005/07/15/ 1120934404860.html). Finally, several papers describe using electronic signage on highways to display both highway and transit travel times. Successful operation of the [changeable message sign] CMS system to disseminate highway driving time in the Bay Area led to the idea of displaying transit information along with freeway travel time. By displaying both travel times, by freeway and by train, travelers can make informed decisions about their commute. Caltrans District 4 is now comparing driving times with riding Caltrain Baby Bullet trains along the US-101 corridor. There are three signs designated to display highway travel time along with transit trip time—to San Francisco and San Jose— for Millbrae and Redwood City stations. These signs are located approximately half a mile from the freeway exit for the nearest Caltrain station. The signs display traveler information when the station-to-station train trip time is shorter than the highway travel time. This feature is designed to encourage motorists to use public transit during rush hours and reduce highway congestion. (18) (Figure 11) FIGURE 11 Caltrans changeable message sign (17, p. 4). Figure 12 shows the effect of the CMS. The results of the study are as follows: • “The user satisfaction analysis indicated positive support for the transit-related CMSs and the general objective of sharing travel information with commuters. • Drivers are likely to change to transit if transit can offer travel time savings over 15 minutes. Due to congestion in the Bay Area, taking transit can be faster during rush hours. In order to make transit more accessible, better information on the transit trip should be provided in addition to the travel time comparisons on the CMSs. Additional information can include real-time parking availability at stations and wayfinding kiosks at the destination station. • A stronger mode switching effect can be expected in the afternoon peak hours due to greater travel time savings with transit. • A network-wide deployment can have large effect on commuting behavior.” (19) FIGURE 12 Willingness to take transit as a result of CMS Transit and Traffic Travel Time Information (18, p. 9). Agencies represented on the oversight panel for this Syn- thesis expressed one primary concern regarding the applica- tion of this type of system: the ability to provide parking structures to support the number of individuals that might exit the freeway and take transit. In the Boston area, the Massachusetts Department of Transportation (MassDOT) and the Massachusetts Bay Transportation Authority (MBTA) are sponsoring a DMS along Interstate 93 southbound that displays the time that the next commuter rail train departs from a commuter rail station located right off I-93: [A DMS] message board positioned just after the I-93 Concord Street Exit (Exit 39) will display the time of the next scheduled train departing from the Anderson- Woburn Regional Transportation Center on the Lowell Line during the morning commute. The message will be displayed [see Figure 13] until several minutes before the next departure to allow customers enough time to exit I-93 and drive safely to the rail station. (20)

15 FIGURE 13 I-93 southbound VMS displaying next train departure time (19). Information Characteristics Literature covering the characteristics of the information dis- played on the signage is limited. In terms of message types and content, the following was offered by Katrin Dziekan: Real-time, at-stop displays must at their most basic level show the numbers or names of PT lines and routes, their directions, and departure times. These constitute specific stipulated user needs. Additional information that is valuable to the customers include: seat availability, arrival time for the next bus or train, and service attributes—such as low-step-up height. The more advanced the display is, the more and better become the potential planning options available to the users, especially during service disruptions. Digital countdown information is clear, and preferred by most (approximately 90%) of users. To differentiate between real-time and static timetable information, it could be wise to implement a standard that follows the natural feelings of the customers—with planned departure time always shown in digital time (for example 10:46), and real-time digital departure countdowns (for example 2 min). (21) In London in 2009, TfL conducted an in-depth evaluation of Countdown signage in terms of what information should be dis- played and the relative importance of each type of information (22). The key findings of the evaluation include the following: • The decision-making process and, therefore, information requirements are very different depending on scenario, ranging from normal service to severe disruption, and high frequency to low frequency services • The simplicity of information provision is key –don’t over deliver. (22, p. 3) In the normal service/high frequency situation, the basic information requirement is to provide reassurance; that is, bus will come in xx minutes. Additional information that is desired: • Advance notice of buses not stopping • Planned engineering works/service cancellation information for all services from that bus stop • Delays/changes to service to onward journey and • Reassurance that more night buses to follow. (22, p. 3) For normal, daily service, the results of this assessment are shown in Figures 14–17. FIGURE 14 Countdown assessment for normal service (21, p. 20). FIGURE 15 Countdown assessment for normal service (21, p. 21).

16 FIGURE 16 Countdown assessment for normal service (21, p. 22). FIGURE 17 Countdown assessment for normal service (21, p. 23). In the minor delays/medium frequency situation, the “information requirement is to provide reassurance and infor- mation to assist if plans need to change, i.e., bus will come in xx minutes but onward delay will be xx minutes. Additional information that is desired includes when an empty bus will arrive (not how crowded the bus is)” (22, p. 4). In the severe delays/low frequency situation, the “infor- mation requirement is to provide information to replan the total journey, so delivery needs to be interactive to be able to provide tailored and specific advice. Additional informa- tion that is desired includes the reason for the disruption and the service status of other modes. At quieter stops, there is a greater need for reassurance as wait times can be much longer and a greater emphasis on alternative bus options in disruption scenarios” (22, p. 4). The recommendations of the evaluation are as follows: • The real challenge is providing appropriate information at the right time, through the most accessible channel • Any information provision needs to be simple, clear and usable • There is a potential danger of over-delivering which would not delight but just confuse • Simplicity would delight especially in the most simplistic of decision-making scenarios • Over-delivery would annoy −‘who cares that the driver didn’t turn up for work −where is my bus?’ • There is no difference in information provision at quieter stops since the decision-making process is identical, only greater reassurance is required. (22, p. 44) As of July 9, 2012, 2,496 Countdown signs have been installed as part of the iBus system (23). U.K. Network Rail has an OIS Process Guide that shows the content of many different types of passenger information displayed on electronic signs (3). The Operational Information System (OIS) is a Network Rail system which is deployed nationally—large electronic display screens are installed at key stations, control centres and some Network Rail corporate offices. The primary objective is to increase passenger awareness of changes to train services either in advance, or in real time due to service disruption. The types of information displayed on the screens include: i. Standard passenger advice (e.g., unattended items and train doors close prior to departure time, CCTV is fitted / Help Points are available, where fitted) ii. Current service information and apology messages iii. General station information iv. Details of planned engineering works v. Rainbow boards (see Figure 18 for an example) (network status’ for TfL and [Train Operating Companies] TOCs, where agreed) vi. Train service performance figures at stations agreed between Network Rail and TOCs vii. Other approved content. (3) FIGURE 18 “Rainbow Board” example (Courtesy: TfL, http:// www.tfl.gov.uk/).

17 to a range of customers with a wide variety of needs. (3, p. 12) “Further, OIS has defined message priority levels within the system so that messages of certain priority take prefer- ence over lower priority messages (in their frequency and also removal of certain lower priority messages). These pri- ority levels are defined” (3, p. 21) in Table 1. The U.K. survey conducted by RTIG (6, pp. 50–53), investigated what channels are being used for disseminating disruption information, approximately how long it takes to distribute information about unplanned disruptions, and the capability of [local authorities] LAs to disseminate information out of working hours. (6, p. 50) During times of disruption, LAs either put a standard holding message on their on-street signs, or give out real time information about the disruption on their signs. Only a few—8 out of 53—actually turn their signs off. Disruption information is provided through a number of channels and most LAs provide information through more than one. 43 LAs across the UK reported using on-street signs to distribute information about unplanned disruptions. 6 LAs reported using only on-street signs for information provision, while 37 use them alongside websites. Most LAs are able to provide information in 30 minutes or less while a few are taking more than 3 hours. Slightly less than half of LAs (23) Various types of information are provided in National Rail stations as follows (3, p. 4): • National Rail Enquiries provides planned engineering works information; • TfL provides London Underground service informa- tion at National Rail stations that are equipped with OIS and are within a 35-minute journey time from a central London station. Also, all London airport railway stations that are equipped with OIS display London Underground service information from TfL. Network Rail identifies “Core Messages,” which— must contain three distinct pieces of information: The Problem What has occurred The Impact What impact will this have on passenger journeys (including any available time estimates) The Advice What passengers should do The over-riding principle of a Core Message must be that it is written as if you were talking directly to the customer (write the message in plain English so that it is easily understood, without using railway terminology or jargon). It must provide information that is of relevance TABLE 1 NETWORK RAIL MESSAGE PRIORITY Priority Level Frequency Base Core Content (not available to OIS operators) Standard passenger advice animations (e.g., unattended items and train doors close prior to departure time) Approved content requests Same frequency as per Low priority messages Will be removed by a Medium, High, or Emergency priority message Engineering Work XML (not available to OIS operators) Planned engineering works XML Same frequency as per Low priority messages Content will be removed by High or Emergency priority message Rainbow XML (not available to OIS Operators) Rainbow Board XML feeds RTPPM information (where agreed) Same frequency as per Low priority messages Content will be removed only by an Emergency priority message Low Restoration of normal services following disruption messages Apology messages Messages advising of additional trains on some routes General passenger advice messages Content will be removed by High or Emergency priority message Medium Non-CSL2 service disruption (line problems with delays above 10 min, except when CSL2 has been declared and active), planned industrial action, emergency engineering works, safety messages relating to weather warnings, irregular congestion due to sporting event/concert, etc. Contingency timetable introduced for entire network and trains running well against plan Displays at twice the frequency as Low and equal frequency to other Medium messages High CSL2 declared and active Contingency timetable introduced for entire network and trains disrupted Do-not-travel messages Displays at twice the frequency as Medium and equal frequency to other High messages Emergency Emergency messages only (e.g., please evacuate the station immediately) Removes all other content.

18 It is important to note that U.S. federal and state regulations and standards are somewhat general in the area of information and communications accessibility. In some sense, technology advancements are ahead of the regulatory process. In addition to meeting federal and state requirements, real-time public transportation information systems should be fully usable by all riders, including riders with disabilities. (24, p. 2) In the United States, several transit agencies have deployed electronic signage that provides audio in a variety of ways: There are multiple approaches to providing audio information. Announcements of DMS displays could be made at acceptable intervals (e.g., every three minutes). Second, a push-button that is pressed and provides an audio announcement of what is displayed on the DMS is another possible alternative. Persons who are blind or visually- impaired would have to be directed to the location of this type of push-button, often done by audio alarm. Third, providing detailed information via telephone is another possible alternative. Finally, an infrared device (e.g., Talking Signs®) that provides the audio equivalent of what is displayed on a DMS could be utilized. This alternative may require that the public transport authority provide visually-impaired riders with the infrared device. Other alternatives, such as a stationary device (e.g., a telephone at a stop/station that is directly connected to the customer information department) that provides direct access to an interactive voice response (IVR) system, should be explored. (24, p. 4) Remote Infrared Audible Signage (RIAS) or ‘talking signs’ provide a signage system for blind, visually impaired, or cognitively or developmentally disabled transit users. RIAS consists of infrared transmitters that continuously broadcast directional information and spoken messages to wireless receivers carried by a user. The handheld devices relay station navigation or traveler information to a user via audio messages. (5, p. 38) RIAS has been deployed in a number of U.S. cities, at select transportation centers and buildings, as well as internationally in Canada, Italy, Japan, Norway, Scotland and Turkey. RIAS has not yet become a widespread proven system. However, some of the U.S. transit agencies do currently operate RIAS technology in selected transit stations. These agencies include the Bay Area Rapid Transit (Powell Street Station and Fremont Station), San Francisco Municipal Railroad (selected stops), and Capital Area Transit Authority in Lansing, Michigan (on all buses). These agencies operate RIAS in support of a single mode of transportation (bus or rail). (5, pp. 38–39) A demonstration of RIAS in the Puget Sound area is the first multimodal application that seeks to provide a seamless connection of signage among different modes. Potential challenges associated with the widespread deployment of RIAS systems include geographic scope and service population, which both impact the associated costs and benefits of system implementation. The effectiveness of a RIAS network is dependent on its comprehensive nature and on its ability to communicate seamlessly inside and outside of transit systems. RIAS networks are optimally provided in combination with other common public signage such as crosswalk and street signals and other directional aids. In addition, the potential size of the visually impaired or cognitively disabled community benefiting from a RIAS network must be commensurate with the associated system deployment costs. (5, p. 39) who responded were able to provide disruption information out of hours, while 32 could not. (6, p. 51) INFORMATION ACCESSIBILITY In terms of accessibility, the literature suggests that there is a growing trend toward providing both audio and visual announcements. Currently, there are no U.S. laws that explicitly address the accessibility of this type of information, which is disseminated via media such as dynamic message signs (DMS), mobile telephones and interactive voice response (IVR) systems. The key issue regarding the accessibility of real-time information is providing the information in alternate formats so that persons with disabilities can access it in an equivalent way to persons without disabilities. For information provided visually, the audio equivalent of that information should be provided. And vice versa, if audio information is provided, the information should be provided visually. (24, pp. 1–2) In fact, there is no specific portion of U.S. law that absolutely states that visual information must be provided in audio format and vice versa. However, several portions of the Americans with Disabilities Act (ADA) of 1990, the ADA Accessibility Guidelines (ADAAG), and rules and regulations issued by some U.S. states (e.g., Massachusetts Architectural Access Board (MAAB) Rules and Regulations) state in various ways that alternate formats must be provided. (24, p. 2) Specifically, this includes the following citations: • ADA: › 49 Code of Federal Regulations (CFR) Part 37; §37.167 Other Service Requirements (f) › 49 CFR Part 37; §37.5 Nondiscrimination (f) › 28 CFR Part 36; §36.302 Modifications in policies, practices, or procedures (b) Specialties (1) › 28 CFR Part 36; §36.303 Auxiliary aids and services • ADAAG: › Sections 218 and 810 of the ADAAG › 4.30 Signage › 10. Transportation Facilities • Electronic and Information Technology Standards detailed in Section 508 of the Rehabilitation Act [4]. While Section 508 requirements are technically only applicable to federal agencies and their contractors, these standards have been widely and proactively accepted by public and private transportation agencies. (24, p. 2) At the State level, there may be requirements that govern the need to provide alternate formats, such as Section 18 of the Massachusetts Architectural Access Board regulations (521 CMR). In this case, Section 18.11— Announcements in Seating and Platform Areas states that ‘Visual systems for providing announcements to deaf and hard of hearing customers shall be provided wherever there are auditory systems for providing announcements.’ In addition to federal and state regulations, standards and guidelines, commitments made at the local level by a public transport agency regarding information and communications accessibility should be considered.

19 In the United Kingdom, RTIG reports that the use of audio at stops that have an electronic sign has grown over the past several years, but has not grown much over the past year. Figure 19 shows various methods that are used to pro- vide audio at stops in the United Kingdom to augment elec- tronic displays (6, p. 25). FIGURE 19 Audio provision at stops 2005–2013 (6, p. 26). In March 2012, RTIG published guidance regarding meeting the needs of disabled passengers for real-time information (25). This document includes specific recom- mendations related to electronic signage in the following categories: • Font and format—Recommendations are provided to ensure clarity and legibility specifically for users with visual and cognitive impairments; • Refreshed text—Recommendations are made regard- ing scrolling or refreshing text where text is too long to fit on a single line on a screen. The recommenda- tions reflect a “tradeoff between the informativeness of the message and its length: a longer message may give more information, but it may be more complex, require scrolling or take more time” (25, p. 7). • Scrolling text—Recommendations are made regarding increasing legibility when scrolling. • Sign finish, contrast and borders—Recommendations regarding these items are made because they “can affect users’ ability to discriminate the information on the sign” (25, p. 9). Further, there are guidelines that state that: – Signs need to be made of materials which do not cause undue reflection or glare – The message should contrast with its background to ensure clarity and legibility – It is important that the borders contrast with the col- ors and materials behind the sign so that the sign is immediately visible. • Sign positioning, lighting and environment on buses— Recommendations are provided to ensure that signs in buses are positioned so that a passenger is not required to search to find them. In general, recommendations include the following: – “Signs should face the largest number of passengers possible – Ideally more than one sign should be provided – Signs need to be well lit, but should not be posi- tioned such that they cause glare – Uniformity of illumination and contrast is impor- tant to those with visual impairments” (25, p. 10). • Sign positioning and environment at stops and shel- ters—Recommendations in this area include signs hav- ing as long sight lines as possible and the potential need to be angled. Further, recommendations include con- sidering providing signs at two heights for standing and wheelchair users. Finally, recommendations discuss that signs should be well lit, but protected from direct sun- light or artificial light to avoid glare (25, p. 11). This report also provides guidelines for audible assis- tance systems in the following categories: • Provision of audible information, messages and announcements—recommendations are made in this covering the following (25, p. 19): – Visual information should be presented in audible form where possible – All audio information should be provided with enough time for passengers to act on the informa- tion and where possible should be repeated – Complete one message before starting another, if possible – Ensuring voice intelligibility so that messages are understood. • Audio systems—Recommendations cover consideration of background noise and having the audio be loud enough to be heard above it, but not so loud as to cause a nuisance. • Hearing enhancement—Recommendations cover sys- tems such as induction loops, infrared or radio trans- mission to help make audio announcements more accessible to hard of hearing passengers by mitigating the effects of distance from the sounds source, ambient noise and reverberation. • Triggering audio assistance with keyfobs (small hand- held device which triggers an audio announcement by means of short range radio waves at a distance of between 5 and 8 meters). • Triggering audio assistance with smartcards. • Synthesized speech—Recommendations cover vari- ous technical approaches to the speech synthesis. In the La Défense subway station in Paris, a touchscreen monitor, which is accessible to persons with disabilities, provides information on public transport, station services, shops, and the station exterior by means of a 40-in. interface (15, p. 111). The touchscreen is activated by waving a hand and then provides “four functions: search for destination by name, search for destination by theme, access latest infor- mation on destinations and surfing the whole map. Both the

20 screen and its cradle are adapted for use by wheelchair users and the visually impaired. ‘The ergonomics of the cradle mean wheelchair users can access the screen face on, by slid- ing their wheelchair underneath. The blind can also locate it using their cane, plus the keys are visually enhanced. The programme itself gives details on accessibility and lifts in the station and its surroundings” (15, p. 112). ACCURACY AND RELIABILITY In terms of accuracy and reliability, Dziekan (21) stated that “real-time information displays must work reliably, other- wise users lose confidence in the system very quickly, as system performance can be checked on the spot. Once that happens, it becomes much harder to satisfy the user, and overcome prior negative experience” (21, p. 19). In London, the iBus system, which replaced the original Countdown system, includes Real Time Passenger Infor- mation (RTPI) at bus stops—“approximately 2000 signs installed at bus stops across London providing estimated times of arrival for vehicles” (26, p. 2). “More reliable equip- ment and an improved prediction algorithm are expected to produce fewer non-predicted buses that fail to show a predic- tion, greater accuracy of arrival times, and for the first time pre-trip predictions i.e. predictions displayed for a bus whilst the vehicle is on a preceding trip” (26, p. 3). MONITORING Several papers reported monitoring the accuracy and reli- ability of the information presented on electronic signage. At the Chattanooga Area Regional Transportation Authority (CARTA), the DMS “were tested for operability, accuracy and reliability during the test period. Additionally, the DMS system was evaluated by the CARTA Technology Director to ensure proper integration with CARTA’s overall ITS sys- tem (27). In Busan, Korea, “The information and the accuracy of the bus information system were found very satisfactory. Satisfaction with the bus headway improved compared with the before-study result by about 9%. The accuracy of bus headway or arrival information and the satisfaction with bus information booths improved, leading to higher reliability of bus arrival information” (28). In Thessaloniki, Greece, “The analysis performed on the data collected from the survey of both regular and cir- cumstantial PT users in the city of Thessaloniki shows that the existing RTPI system is generally evaluated positively. Satisfaction levels are quite high, more than 80% for both the content and the reliability of the information given” (11, p. 254). STANDARDS Several papers discussed the current standards for the implementation of electronic signage. One underlying stan- dard that is being deployed extensively in Europe and now in the United States is SIRI. “SIRI specifies a European inter- face standard for exchanging information about the planned, current or projected performance of real-time public trans- port operations between different computer systems” (29). “SIRI is intended to be used to exchange information between servers containing real-time public transport vehi- cle or journey time data. These include the control centres of transport operators and information systems that utilise real-time vehicle information to operate the system, and the downstream systems that deliver travel information to the public over stop and onboard displays, mobile devices, etc. SIRI uses eXtensible Markup Language (XML) to define its messages” (29, p. 3) “SIRI uses open standards and is plat- form independent in that it is free to use and can be deployed onto any general computer operating system that supports XML” (30). Two of the services available in SIRI are the “Stop Timetable and Stop Monitoring services. The Stop Timeta- ble (ST) and Stop Monitoring services (SM) provide stop- centric information about current and forthcoming vehicle arrivals and departures at a nominated stop or Monitoring Point, typically for departures within the next 20–60 min- utes for display to the public. The SM service is suited in particular for providing departure boards on all forms of device” (29, p. 7). Figure 20 shows how this service works with DMS. SIRI provides the following eight services: • Production Timetable Service • Estimated Timetable Service • Stop Timetable Service • Stop Monitoring Service • Vehicle Monitoring Service • Connection Timetable Service • Connection Monitoring Service • General Messaging Service TransXChange (TxC) is another standard used in the implementation of DMS. “TxC is an XML based UK stan- dard for exchanging PT data. It is used by the Traffic Area Networks (TAN) and the Vehicle and Operators Services Agency (VOSA) to register bus schedules electronically. It is also used to exchange such PT information with other computing systems such as journey planners and RTI sys- tems” (30, p. 15). TxC has five services: Creating Schedules, Submitting Schedules, Publishing Schedules, Exchanging Schedules and Importing Schedules (29, p. 16). Although the word “schedule” is used in these service titles, real-time information is included in these services.

21 The Virginia Department of Rail and Public Transportation (DRPT) has led an effort to create a technology community for transit operators statewide and has commissioned a standards working group. DRPT is interested in making real-time and historical data available to the public and to 3rd party developers in order to improve passenger information, government transparency and multimodal transportation options in the state and is using the working group to ensure statewide coordination. (33) The Transit Real-Time Traveler Information Standards Working Group examined several relevant standards, including “GTFS, TCIP—transit communications interface profiles, SAE J-2354—defines multimodal traveler itinerary requests and responses, and SIRI” (33, p. 1). REQUIRED RESOURCES Several pieces of literature discuss the resources required to implement, operate, and maintain electronic signage. Ferris et al. (34) state that it is likely to be prohibitively expensive to provide and maintain such displays at (for example) every bus stop in a region. With the increased availability of powerful mobile devices and the public availability of transit schedule data in machine readable formats, there have been a significant number of tools developed to improve the usability of public transit, especially mobile tools. One motivation is that, as noted above, it is unlikely that real-time transit information will be available on a public display at every stop. Another is that personal mobile devices can also support additional, personalized functionality, such as customized alerts. As mentioned earlier, the display system at Southampton Airport uses standard modern Internet processes to transfer the information from the data providers to the travel display system. Essentially, the system acts as a real-time travel information nexus, automatically and continuously gathering data from remote data sources of travel information using standard communications protocols such as XML (eXtensible Markup Language) and Internet protocols such as SOAP (Simple Object Access Protocol). This data is then automatically formatted into a defined standard XML format by our adapter software written in Java. (8, p. 3) In Johannesburg, South Africa, the BRT “DMS signs sup- port the [National Transportation Communications for ITS Protocol] NTCIP protocol which provides a standard com- munications interface” (10, p. 6). NTCIP “is a family of stan- dards that provides both the rules for communicating (called protocols) and the vocabulary (called objects) necessary to allow electronic traffic control equipment from different man- ufacturers to operate with each other as a system” (31). A new standard that is being used is to provide real-time information to various applications is called the General Transit Feed Specification (GTFS)-realtime: GTFS-realtime is a feed specification that allows public transportation agencies to provide realtime updates about their fleet to application developers. It is an extension to GTFS (General Transit Feed Specification), an open data format for public transportation schedules and associated geographic information. GTFS-realtime was designed around ease of implementation, good GTFS interoperability and a focus on passenger information. (32) FIGURE 20 SIRI stop timetable (ST) and stop monitoring services (28, p. 7). AVMS = automatic vehicle monitoring system; ITCS = Intermodal Transport Control System.

22 Further, the investment in website and phone-based real time transit information can also save an agency substantially in deployment costs. As an example, Portland deployed their Transit Tracker program in 2001 with information displays at stops, a webpage and more recently a phone system. The transit tracker signs at light rail stations and 13 bus stops in Portland cost $950,000 including message signs and conduit. The cost for computer servers and web page development was much cheaper at $125,000. Given the widespread availability of cell phones and web access, providing real time transit information via a service such as OneBusAway [http://www.onebusaway. org/] could yield a substantial savings for an agency over constructing real-time arrival display signs. (35) As mentioned earlier, the price of the two-screen displays deployed in some locations in Belgium (€17,000) caused the National Railway Company of Belgium (SNCB) to imple- ment a digital advertising network including LCD screens with portrait orientation and panoramic screens that display both advertising and traveler information (15, pp. 113–114). In Thessaloniki, Greece, “the overall acceptance level is high and it seems that the implemented system is both eco- nomically feasible and financially viable. In economic terms annual benefits are twice the investment cost of the system. The main financial benefits come from the reduction of the PT operations costs as a result of the AVL system” (11, p. 254). DECISION PROCESSES A variety of literature describes the processes that agencies use to determine if electronic signage should be deployed and where it could be located. First, literature describing the ben- efits of electronic signage was reviewed because it was found that the identification of benefits often contributes to the deci- sion about whether signage will be deployed. Several papers discussed how real-time information often reduces passen- gers’ perceived waiting times at stops and stations: “Although stop signage with next-arrival information does not directly reduce wait times, since passengers have to be at the stop to know this information, it reduces anxiety and may provide a perceived benefit of less safety and security risk. By knowing next-bus arrival information, passengers may be able to make better use of their time or seek alternate modes of transporta- tion (e.g., if the wait time is too long)” (36). The effect of providing real-time information displays at transit stops has been reported in several locations around the world, as follows (37): • In Stockholm, a study showed that passengers without real-time information displays at their stop overesti- mated their wait time by 24% to 30%, compared with those who had real-time information at their stop, who overestimated their wait time by 9% to 13%. • “In London, the provision of real-time information at stops was found to reduce perceived wait time by 26 percent” (37). • In the Netherlands, the introduction of passenger infor- mation displays on a tram line in The Hague resulted in a reduction of perceived wait time by 20%. Generally, “it is well known that people like at-stop real- time information and have very positive attitudes towards it” (38). Dziekan and Kottenhoff also described the most significant benefits of electronic signage at stops: increased feeling of security, reduced uncertainty, increased ease-of- use, increased willingness-to-pay, adjusted travel behavior and other adjusting strategies (such as letting a crowded bus go by if the display showed that another would be arriving shortly, and adjusting walking speeds according to the infor- mation received by at-stop real-time information displays), mode choice, higher customer satisfaction, and better image (38, pp. 492–495). In a 2011 survey of New York Metropolitan Transporta- tion Authority customers, Countdown clocks are having a positive impact on the overall customer experience. All 54 subway service and station attributes were rated higher by those with countdown clocks in their station than those without a countdown clock in station. This includes: • Customer Information: – Knowing how long until next train arrives – Clarity of announcements on station platforms – Information in station about unscheduled delays • Service Performance: – Predictability of subway travel time – Service frequency – Service reliability • Personal Security: Personal security in station after 8 p.m. (39, p. 6) Further, satisfaction with countdown clocks in 2011 was 59% very satisfied and 37% satisfied for a total of 96%. “Sat- isfaction improved statistically on information about delays, reflecting improved planned service change posters and LEDs on countdown clocks” (39, p. 12). Although not specific to electronic signage, the research conducted by Tang and Thakuriah (40) implies the following: • Real-time information systems provision should be considered as one way to improve transit ridership. The psychological benefits brought about by such systems are one of the reasons leading to the ridership gain. • Real-time information systems have the potential to serve as a good intervention to change the travel habit of current transit non-users and increase their transit use, thus such systems have the potential be used as a tool to increase transit mode share. • As past experience with real-time transit information systems has positive effect on commuter’s attitudes and intentions to increase transit use if such a system is

23 marked, rising from 1,972 flag installations in 2010 to 2,872 in 2011. This rise of 900 signs is largely accounted for by apparent growth of 740 in the East Midlands. However, this is the result of a new return which we did not have last year” (6, p. 24). TABLE 2 SIGNAGE LOCATION IN GREAT BRITAIN* Location Position within GB end-2011 No. RTI Displays RTPI Physical Display Location Flag installation Shelter Transport Hub Private building Other Metropoli- tan 4,453 487 2,453 594 41 2,501 Non-Metro- politan 3,532 2,385 2,527 306 95 5,193 London 2,000 - 2,000 - - - TOTAL 9,985 2,872 6,980 900 136 7,694 *“Note that there are more signs accounted for in physical locations than the total number of physical signs (for instance the sum of total shelter, total flag, total hub etc displays is 10,888, which is 904 more than are accounted for by the LEDs and LCDs). We have not included “other” in this calculation since a number of people included virtual signage in this category” (6, p. 24). In Plymouth, United Kingdom (41), the Public Transport Information Strategy includes the continued “rollout of RTPI displays in bus shelters throughout the City but will give pri- ority to stops on the selected Quality Bus Corridors” (41, p. 16). Quality Bus Corridors are defined as important strategic routes that are improved to increase bus use. Improvements are introduced to make buses more reliable and passenger waiting facilities more efficient and comfortable (42). provided, for transit systems that are planning to deploy such systems, the facilitating programs to familiarize commuters with real-time transit information systems may help to increase transit ridership. (40) SELECTION CRITERIA In terms of selecting electronic signage over other dis- semination media, the literature describes various selection processes. In Caulfield and O’Mahony (37), “respondents were asked to choose between three options of accessing real-time public transit stop information: SMS, a passenger information display, or a call center (37, p. 5). The results show respondents derive the greatest benefit from real-time public transit stop information displays. This result was as one would expect, as this is one of the most effective meth- ods of relaying real-time public transit stop information” (37, p. 18). SIGNAGE PLACEMENT In terms of where signage should be located, the literature describing the status of electronic signage in the United Kingdom identifies the locations as shown in Table 2 (6, p. 23). “By the end of 2011, the majority of [real time informa- tion] RTI displays were installed in shelters (64%), or flag [pole-mounted] installations (26%). Overall this marks a shift towards flag installations and away from shelters com- pared with 2010. Signage in all locations has risen since last year, though the increase in flag installations is particularly