Transportation Deployment Casebook/2014/Automatic Vehicle Locators on Fixed Route Bus Systems

Mode Description and Uses
Automatic Vehicle Locators, or AVLs, are communications devices that allow transit agencies to track the location of transit vehicles. Once bus location is known in real-time (or close to real-time), there are a wide variety of uses for this information. The most visible application to the public is to provide service information to users (eg. count down to the arrival of the next bus, delay alerts, etc.) via digital signs or, more recently, websites and smart phone applications. AVLs allow transit agencies to improve bus schedules and service efficiency because they know where all buses are at all times. AVLs are also commonly used to provide location information in an emergency situation, trigger onboard “next stop” announcements, and for signal prioritization.

This page contains information about the life cycle and deployment of AVLs on fixed route bus systems in the United States.

Early AVLs
Early AVLs were deployed by transit agencies for operational purposes. Having the ability to monitor the location of all buses from a central dispatch was a major advance for transit agencies. AVL systems allowed agencies to:
 * improve response time to emergencies
 * develop performance measures based on AVL data
 * observe and improve bus schedule adherence
 * refine scheduling and dispatch
 * coordinate transfers
 * monitor driver performance
 * facilitate on-street service adjustments
 * improve bus stop placement

Transit planners, engineers, and dispatchers today can imagine how different and difficult it was to schedule buses without information generated from AVLs. One dispatcher from King County, Washington described it as “working with blinders on.” Many transit agencies deployed limited AVL systems for specific purposes. For example, the Metropolitan Council Transit Operations in Minneapolis/St. Paul, MN used AVLs to determine when buses departed and arrived from garages or maintenance stations.

Signpost
Most early AVL systems relied on signpost technology. In this system, communication devices on the bus sent radio waves to devices located on the side of the road. As the buses drove along their routes, the devices on the buses could “see” the signposts on the road and communicate information about bus location and time information back to a central information hub.

Inside the bus, the device was composed of an antenna and a transponder. The antenna, located on the roof of the vehicle, received the signal from the signposts on the road side and sent the information to the transponder, which sent the information to the radio and communicated the information to dispatch. The bus’ odometer provided additional information to determine bus location in-between signposts. The opposite system was also used, where signposts detected vehicles and the signposts transmitted the information to the central information hub.



Along the route, signposts were built on tall polls in order to have clear, direct access to the antennae on the top of the bus. Signposts were often placed on utility polls and thus connected to the power grid, but some were battery powered. The two major technologies for Signpost AVLs are demonstrated in Figure 1 to the right. Signposts were either “sharp,” sending a narrow signal band, or “broad,” covering an entire intersection. In some systems, loop detectors were installed in the pavement instead of signposts on the side of the road. When a bus passed over an inductive loop detector, the bus number, time, date, and loop detector number were recorded and sent to dispatch. Location and density of signposts determined the accuracy of this communication technology. Signpost technology was used most often in relatively dense cities with many fixed-route bus systems, and signposts were placed at intersections used by multiple bus routes. Early adopters of Signpost AVLs included Chicago (1969), Toronto (1972), Cincinnati (1975), Los Angeles (1977), and New York (1979).

Ground Based Radio Positioning (Loran-C)
A few early AVL systems relied on Ground-Based Radio Positioning, most notably Loran-C. Loran-C was a long-range navigation system deployed and maintained by the U.S. Coast Guard, which emitted low frequency radio waves. First introduced in 1957, Loran-C radio waves covered the entire U.S. through 2010, although it was used in AVL detection only through the early 1990s.

AVL systems using Loran-C installed Loran-C receivers on transit buses. Each Loran-C transmitter emitted signal at a specific, known frequency. Receivers located on buses detected Loran-C signals arriving from two to six known sites and estimated their location (within an accuracy of 150 ft) based on the differences in arrival time of the Loran-C signals from different transmitters.

Global Positioning Systems (GPS)
The use of LORAN-C and Signpost AVLs declined sharply in the late 1990s with the rise of Global Positioning System (GPS) based AVL systems. Originally designed by the U.S. military, GPS uses satellites to determine the location of a receiver with precision of within 10-20 meters. The U.S. military set the goal of creating a defense navigation satellite system in 1973 and launched the first GPS satellite in 1978. By 1990, the system was functional and was used heavily by the military in the first Gulf War. In 1995, the 24th GPS satellite went into position, allowing the satellites to fully cover the globe and making the system fully functional.



GPS became available for civilian aviation in 1988, fulfilling a commitment made by President Ronald Reagan in September 1983 after the Soviet Union shot down a Korean Airlines Flight 007 that unintentionally strayed into Soviet airspace. Although free for general (non-aviation) civilian use starting in 1996, GPS was not reliable for transportation purposes because the satellites sent out two signals, one for military and one for civilian use, and the U.S. intentionally degraded the civilian signal for security reasons. GPS signals at that time were accurate to 100 meters. In 2000, President Clinton ended this practice, resulting in a significant increase in accuracy and the rapid growth of GPS use as AVLs in transit. Currently, the U.S. Air Force flies and maintains 31 satellites at an altitude of around 12,000 miles in order to ensure that at least 24 satellites are operational so that GPS services can be maintained around the globe.

GPS systems depend on communication with satellites signals, which can be interrupted by tall buildings. As a result, some transit systems use a technique called dead-reckoning, which uses odometer readings to estimate the bus’ location when GPS signal is lost. Dead reckoning was used in most transit GPS systems when the degraded GPS system was in place.

Federal support for early AVLs
Although technology for Signpost AVL systems has been around since World War II, bus systems were slow to take advantage because of the high cost of installing and maintaining signposts. Deployment on U.S. systems didn’t begin until federal agencies began to show an interest and to fund AVLs. All of the early AVL adopters (Chicago, Philadelphia, Los Angeles, Cincinnati) were federally funded. In 1968, the U.S. Department of Housing and Urban Development (HUD) proposed large scale deployment of AVLs in its Report to Congress entitled “Tomorrow’s Transportation,” and funded the Chicago Transit Authority (CTA) to improve public transportation through AVLs. This initial effort installed AVLs on 500 buses and signposts on the night bus routes in Chicago. In 1974, the Urban Mass Transit Administration (or UMTA, later the Federal Transit Administration or FTA) funded testing of AVLs in Philadelphia, followed by Los Angeles and Cincinnati, Ohio. The purpose of these tests was to assess the role and accuracy of AVLs in schedule adherence. In 1977, UMTA selected the Southern California Rapid Transit District to showcase the benefits of AVLs.

Federal support for early GPS
As mentioned above, the U.S. military is responsible for the invention, deployment and maintenance of GPS satellites. Presidents from Reagan to Obama have supported and expanded policies and funding for GPS for both military and civilian uses. Most recently, President Obama’s National Space Policy states: “The United States must maintain its leadership in the service, provision, and use of global navigation satellite systems” (2010). The U.S. government continues to provide open and free access to GPS satellites and to the information needed to build systems like APL that depend on GPS.

GPS is funded out of general U.S. tax revenues. Most of the program funds come through the Department of Defense, which is responsible for operations, maintenance, and upgrades to the GPS system. Additional funding comes from the Department of Transportation. In 2014, Congress allocated about $1.2 billion to fund the GPS program, about $55 million of which came from the Department of Transportation.

GPS is used in many fields other than transportation, including construction, farming, mining, shipping, emergency response, weather forecasting, and even synchronizing communication networks and power grids. However, it is interesting to note that it is the U.S. Department of Transportation, not the Department of Agriculture or Energy or the Interior, that has been assigned responsibility for funding civilian upgrades to GPS. This highlights the importance of GPS systems to transportation operations, including transit system operations but also systems like highway operations, maintenance, and tolls.

The Federal Transit Administration (FTA), an agency within the Department of Transportation, continues to fund AVL systems in transit agencies, but this funding is sporadic. For example in 2009, through the American Recovery and Reinvestment Act (ARRA), FTA provided $2.8 million dollars to the City of Santa Clarita, California, $7.7 million to Lehigh and Northampton Transportation Authority in Allentown, Pennsylvania, and $4.3 million to the City of Durham, North Carolina.

Growth and Decline of Early AVLs
AVLs offered many advantages to transit operators, but despite technology being readily available, deployment of signpost AVLs was gradual and never achieved more than 60% deployment. Signpost AVLs faced a number of challenges. The first challenge was the cost. Transit agencies have been strapped for cash since even before the 1960s, and AVL systems required significant investment. Federal funding played a significant role in deployment prior to 1995. Deployment of AVLs often required additional fleet management software and communications upgrades, which increased the price tag significantly. For example, total system costs for AVL systems were $7 million for MARTA in Atlanta, $6.6 million for Tri-Met in Portland, and $31 million for NJ transit (in 1997 dollars). These costs don’t include staff training, which was an additional expense.

Another challenge was system maintenance. This was especially burdensome with the Signpost AVLs (as compared to the current GPS system) because each individual signpost had to be maintained. Depending on the system, this required maintenance of hundreds or thousands of signposts spread across an urban area.

Additionally, because AVLs were new, the information they offered was not expected by consumers, and transit agencies found it hard to justify a new system that was not in demand and that was not (at that time) known to increase ridership. Over time, however, agencies reported significant operational (and cost-saving) advantages to AVL systems including. For example:
 * schedule adherence improved by 25% in Baltimore
 * manual schedule adherence checking could be eliminated, resulting in cost savings
 * customer complaints could be verified (in Milwaukee, they found that 50% of complaints were invalid)

Demand for these systems became more pronounced because of the ability to provide riders with “real time” schedule updates. A 1994 survey in London after they implemented digital Countdown signs at bus stops found that:
 * 89% of passengers felt waiting for a bus was more acceptable with real-time information displayed
 * 83% felt that time seemed to pass more quickly when they knew how long they would have to wait
 * Waiting at night was perceived as safer.

Despite the costs, more and more transit agencies saw the operational and customer service benefits of AVLs, and system deployment went through a growth phase between 1975 and 1988. Early AVLs peaked in 1991, when about 60% of bus systems in the United States had deployed this system.

It is likely that deployment would have continued if not for the fact that GPS technology became usable for civilian purposes around that time. GPS has the significant advantage that it does not require installation of sensors all around the network. Transit agencies who did not yet have AVLs chose GPS systems rather than signpost systems, and those with signpost AVLs gradually switched over as well. Signpost AVLs were phased out by the early 2000s.

Growth of GPS-based AVLs
Once GPS technology was available to transit agencies, it became the logical choice for AVL systems because it was much simpler to deploy. Buses had to be equipped with GPS receivers, but the satellites were maintained by the U.S. military and the signals were free and open to use. Early adopters of GPS AVLs include Denver (1996), Portland’s Tri-Met (1997), New Orleans’ RTA (1997), Chicago Transit Authority (1999), and Baltimore’s MTA (1997).

GPS AVL systems continue to be a major capital cost for transit agencies. For example, in 2002, San Francisco MUNI equipped 827 vehicles with GPS for $9.6 million dollars. Portland’s Tri-Met equipped 689 buses for $7 million.

Operational benefits of GPS
Due to parallel technological advances, GPS systems offer additional benefits beyond those possible with early AVLs. These include:
 * automatic onboard announcements triggered by location as the bus approaches the next stop (helps agencies comply with ADA requirements)
 * fare box (smart card) and ridership information linked to GPS data
 * mechanical and electronic vehicle monitoring systems linked to GPS in order to monitor vehicle status
 * hidden alarm buttons to send emergency calls to dispatch, complete with location information
 * improved transit signal priority systems based on exact location, speed, and schedule adherence
 * video surveillance
 * payroll

The increasing popularity of AVLs is due in part to the growing understanding that these systems were worth the cost because they would pay off in terms of ridership and improved operational efficiency. In one study, Denver, CO, and Milwaukee, WI reported 5% increases in ridership attributable to adjustments in headways and run times due to AVLs.

Customer satisfaction
A continuing impetus behind the growth of AVLs was that transit agencies found almost universally that providing customers with real-time schedule updates had a significant positive influence on customer satisfaction. Perception of bus reliability increased even more than actual schedule adherence when customers were provided with real-time information about when the bus was coming. Transit agencies reported notable decreases in customer complaints following AVL deployment; in the year immediately following AVL deployment, customer complaints fell by 24% in Milwaukee, WI, 26% in Denver, CO, and 53% in Portland, OR.

In the 1990s and early 2000s, the most common way to provide bus riders with AVL information was through digital signs posted at bus stops. These digital, or “dynamic message” signs, provided customers with current time and date, route number, destination of the bus, wait time until the next bus, and service interruption or security messages. Some systems introduced automated telephone services that provided passengers with real-time bus arrival information. For example, Denver, CO’s “Talk-n-Ride” provided information to passengers on the next three buses at any given stop.





Beginning in the early 2000s, agencies began posting real-time bus arrival information on the internet. Depending on the system, this included maps showing bus location, applications that would provide directions using real-time bus information, and/or simple “next bus” information.

With the proliferation of smart phones, real-time bus information has moved onto smart phones, which detect a person’s location and provide real-time transit information. While this information can be accessed through websites on smart phones, many agencies (or, often, third party smart phone application developers) now provide more user-friendly smart phone applications with the same information. These services have proliferated on transit systems with GPS AVL information, from NextBus, San Francisco MUNI’s mobile website, to OMG Transit, a third party application in Minneapolis- St. Paul, to BusTracker in Tulsa, OK, and MyBus for NJ Transit.

Future directions
In 2013, over 70% of fixed bus transit services in the U.S. had deployed GPS AVLs. Given the extensive and growing list of uses for GPS AVLs, it seems highly likely that the system will continue to maturity and to 100% deployment. Not only have transit agencies come to depend on AVLs for system planning and operations, but customers have come to expect real-time bus information, and transit agencies now have to justify lack of AVLs rather than presence. This is especially true for choice riders, who may not ride a system that they find unpredictable or unreliable.

The U.S. federal government is committed to maintaining and modernizing GPS technology, and it continues to launch GPS satellites of increasing technological sophistication. Improving integration of “intelligent transportation technologies” such as AVL is a stated priority of the U.S. Department of Transportation; FTA expects that improved customer service due to AVLs may help increase transit ridership, although there is not a lot of federal funding available for such investments. Local transit agencies are likely to invest in AVLs as is financially feasible and GPS systems are likely to take on an ever increasing role in transit system planning, operations, and customer service.

Data
Data for this analysis came from four major papers, some from the FTA (Federal Transit Administration) and some from the American Public Transit Association (APTA). Information on deployment of the first AVLs (before 1991) came from the FTA’s Transit Cooperative Research Program (TCRP) Synthesis 24 which, published in 1997, provided a brief deployment history of AVLs for bus transit. Information on deployment of early AVLs from 1991 through 1997 came from the FTA report Advanced Public Transportation Systems: The State of the Art 2000, which included deployment information for Signpost, Loran-C, and early GPS systems.

GPS-based AVLs came from two sources. The first was the APTA’s 2011 Public Transportation Fact Book Appendix A: Historical Tables which provided GPS deployment information for a sample of agencies in the 2010 APTA Public Transportation Vehicle Database. The other was from a 2011 study conducted by the FTA’s Research and Innovative Technology Administration (RITA) which surveyed 158 transit agencies in the U.S.

It is important to note that this data is not comprehensive, but is based on AVL deployment information of a sample of transit agencies that were respondents in different surveys. Comprehensive AVL deployment information was not available, but the data provided in the studies cited here is sufficient for a broad lifecycle analysis.

Regression Analysis
In order to better understand the life-cycle of AVLs, I conducted a single variable linear regression analysis on AVL deployment. The early AVLs were treated as a separate curve from the GPS AVL systems because although the systems overlapped in time, they were deployed separately.

For the regression, the dependent variable (Y) was the % of transit systems with AVL systems in use on their fixed-route bus systems. The independent variable (X) was the year.

Y = ln (% of systems with AVL) / (K-% of systems with AVL)

K is the estimated maximum deployment at maturity. Because data was not comprehensive and was available only as percentages of transit systems with AVL deployed, K was known. For early AVLs, K was 60% (the level at which signpost and Loran-C systems peaked). For GPS AVLs, K is expected to be 100%.

Results for Signpost and Loran-C
Signpost and Loran-C AVLs have already completed their lifecycle and have been phased out. The growth and decline curves are shown to the right; and Tables 1 and 2 show the results of the regression.

From the growth curve, it is clear that the lifecycle of early AVLs was cut short by the advent of GPS systems. In the growth curve, actual deployment exceeds predicted deployment, but then it abruptly changes direction and begins to fall. T0, the inflection point, was in 1982, and the growth phase was between 1977 and 1986.





Results for GPS AVLs
GPS AVLs are still in the growth phase but are nearing maturity. The deployment curves from two different studies of AVL deployment, along with predicted regression curves for each, are posted to the right. Tables 3 and 4 show the results of the regression. Although the two data sources result in similar (but not identical) graphs and regressions, I provided the results from both studies to illustrate the uncertainty associated with using sample data. Sample data provides an approximation of the real situation.

Looking at both of these curves suggests that GPS AVLs went directly into the growth phase, essentially skipping the birth phase. This is logical since signpost and Loran-C AVLs had made clear the advantages of AVLs, so transit agencies were already thinking about them. The inflection points for the two different data sets were in 2004 (RITA) and 2007 (APTA). The growth phase started in about 2000 and is predicted to level off and reach 100% deployment around 2017. 2007