DC Metro : Getting Back on Track

One year ago, in March 2016, the entire Washington, DC, subway system was closed for 29 hours for emergency inspections. This shutdown came after a number of electrical fires oin the subway system, involving fraying third-rail electrical cables. In January 2015, a Washington Metro train encountered heavy smoke near the L’Enfant station due to a third-rail electrical issue and was forced to cease service. One passenger died from smoke inhalation and others were injured. On March 14, 2016, an electrical fire, caused by the same electrical issues as the Nee L’Enfant station incident, occurred near another station. There were fortunately no fatalities. Still, the Metro management shut down subway service a few days later to allow for a system-wide inspection of all third-rail power cables to proactively address system safety before further incidents.

Run by the Washington Metropolitan Area Transit Authority (WMATA), Metro is the second-highest use rapid transit system in the United States, behind just  the New York City subway system, in terms of passenger trips, serving over 700,000 riders per weekday.  Metro is just over 40 years old and faces the many of the same challenges as older US transit systems, including inadequate funding and maintenance backlogs.

In May 2016, WMATA introduced SafeTrack, a comprehensive accelerated maintenance and repair program for implementing safety recommendations and needed upgrades to rail infrastructure.  SafeTrack involves the use of “surges,” intensive work on specific sections of the rail network and the shutting down of one or both tracks in those sections during this work, together with the reduction of Metro operating hours at night and on weekends to make more tracks available for maintenance.

Last week, the US Government Accountability Office (GAO) released a report on its audit of the SafeTrack program.  GAO found that WMATA did not following leading management practices and “(1) comprehensively collect and assess data on its assets, (2) analyze alternatives, or (3) develop a project management plan”  prior to implementing SafeTrack.  In response to the GAO findings, Metro General Manager and CEO Paul Weidefeld stated that WMATA didn’t have time for comprehensive data collection before starting SafeTrack, because safety issues and delayed maintenance had reached a critical point and needed to be addressed as soon as possible. GAO recommends that WMATA develop a full asset inventory and a project management plan for those needed projects that may not qualify as major capital projects.  WMATA is now working to address GAO’s recommendations.

The GAO report found that SafeTrack “will require an additional $40 million in fiscal year 2017 funding.” It is not yet clear where that funding will come from.  Although many transit systems are challenged by inadequate funding, Metro is specifically impacted by one funding issue not faced by other large US transit systems:  Metro has no dedicated funding or revenue sources for its operating budget. WMATA relies heavily on year-to-year subsidies from the governments of Virginia, Maryland, and the District of Columbia, which each have budget constraints and funding priorities of their own. In 2016, 47% of Metro’s budget came from local and state subsidies and 45% from fare revenue. In contrast, for the MBTA, 62% of the budget comes from dedicated revenue (such as the sales tax) and 33% from fares.  In New York, the MTA’s budget relies 36% on dedicated funding, 52% on fare revenue, and 8% on local and state subsidies.  WMATA currently has an almost $300 million annual budget gap.


The Federal Transit Authority (FTA) provided some funding for SafeTrack repairs and maintenance. Increasingly, business leaders, DC officials, and others are calling for a dedicated source of funding or regional sales tax surcharge to support Metro operations. So far, these requests have faced opposition from Virginia and Maryland officials.  Proponents argue that dedicated funding is not only important for Metro system safety, but could relieve traffic congestion and spur economic development as well.

Also, last week, board members of Metrolink, the regional rail system in Los Angeles, met with the Metro Board Safety Committee to share Metrolink’s firsthand experience with the importance of making safety a priority.  The Metrolink officials showed a poignant video that Metrolink made following the most deadly crash in Metrolink history, a 2008 crash in which 25 people were killed when a commuter train collided with a freight train.  The video focuses on commitment and responsibilities of the Metrolink board regarding safety.  At the meeting,  Metro board member Michael Goldman suggested Metro could create its own video on the safety in the Metro system for its board members and the public.

By: Tracy Zafian, UMTC Research Fellow


The Phantom Bus Driver: Helsinki Rolls Out Autonomous Public Transit

Helsinki, Finland has long been on the forefront of developing cutting edge transportation technologies. By 2025, they hope to implement a “mobility on demand” system that would eliminate the need for private vehicles through the combination of bicycle-sharing, public transit, and on demand taxi services. One of Finland’s laws is particularly conducive to increasing the technology involved with transportation – they do not legally require vehicles on public roadways to have drivers within the vehicle.


In August, they began taking an even more dramatic step to revolutionizing their citizens’ daily transportation needs. Although autonomous busses have been seen before in more controlled environments such as college campuses, the Helsinki bus is the first of its kind to operate on public roads, interacting with live traffic and having to make complex driving decisions. As of November 1st, the busses are running a route between Tampere University of Technology and Hervantakeskus Shopping Centre. The brains behind the project plan on stopping the service at the first snow fall in order to test the vehicle under difficult conditions. By getting commuters out of private cars and into public transit, the city of Helsinki could decongest streets, creating a safer atmosphere for pedestrians, cyclists and drivers.

Developed by French company EasyMile in collaboration with the Metropolia University of Applied Sciences, the model, EZ10, is able to carry 12 passages, 6 sitting and 6 standing. It uses a system of sensors and software in order to be aware of its surroundings. Passengers can board and disembark at predetermined points along the route.

Although the busses are a large step forward in moving toward autonomous transportation, there are still various pitfalls that must be first overcome. First of all, the busses are not completely autonomous. There is an attendant in the front of the vehicle, ready to push the emergency stop if the situation arises. Furthermore, the busses are only currently running at 7 mph, making efficient travel a bit of a difficulty. Lastly, it is not capable of lateral movement – if the vehicle needs to swerve around an obstacle, the attendant must manually do so.


Currently, the best use for the autonomous bus is in last mile service. The city of Helsinki, along with the University, hope to use the bus to move people from a transportation hub, to a final destination in the home. The city does not plan to replace the entire public transit system with these autonomous vehicles, but rather, hopes to use them as supplements to the existing system in high use areas. The main usage Helsinki has in mind is using them as a feeder service, transporting people to faster, more efficient forms of transit. Although only cruising along at a snail’s pace, Helsinki hopes for the bus to finally reach the Finnish line.


By: Adrian Ayala, UMTC Research

Evolving Strategies for Demand Responsive Transit


For people who are unable to drive or use conventional transit (e.g., fixed route buses and trains), getting around can be a real challenge. One group is receiving increasing attention in the transportation community: people with physical or mental disabilities that prevent them from being able to use existing buses and trains. The Americans with Disabilities Act of 1990 (ADA) requires transit agencies to operate curb-to-curb paratransit with ¾ mile of fixed route bus services for these. Although ADA paratransit constitutes only 1% of transit trips in U.S., the services make up 8% of the operating costs. Furthermore, demand for ADA paratransit increased by 41% from 2000 to 2010, and the trend of increasing demand and increasing cost is expected to continue as the American population gets older [1]. This presents a major challenge for transit agencies: equitable service must be provided for customers with disabilities, but increasing costs threaten the ability of agencies to continue providing adequate ADA paratransit along with conventional services. Recent and ongoing research at UMass Amherst addresses multiple strategies for managing ADA paratransit needs.

One way to approach the problem of mounting paratransit costs is to focus on optimizing the operations. Recent studies of ADA paratransit demand and operation patterns in New Jersey have shown that the total operating cost in a service region can be modeled based on the area of the region, the rate that trips are requested per time, and the allowable time window for an on-time pick-up [2].  There are ways to geographically align service regions to cover large areas in order to minimize the negative effects of making customers transfer.  It can be beneficial to break up large regions into zones such that one zone provides service within a dense urban core, and another zone provides service to more distributed areas [3].

Another approach to the problem is to manage demand by incentivizing users to travel at times of day when there is excess system capacity. The current ADA regulation requires agencies to schedule paratransit service within one hour of the customer’s requested pick-up time and to charge no more than 1.5 times the fare of conventional transit service. Peaks in demand at certain times of day leave agencies with no choice but to purchase more vehicles and hire more drivers, but these resources are costly when they go unused at other times of day. A time-varying fare, within the ADA constraints, could incentivize users with flexible schedules to travel at less costly times of the day to improve the system’s overall efficiency [4].

An emerging question is what role existing ADA paratransit should play in serving this population in the long term. We know that shared-ride services are most efficient in areas with dense demand.  In the suburban fringe, there are many trips that could be served more cost-effectively by taxis or on-demand mobility services (e.g., Uber, Lyft). In the Boston area, where the average cost of serving a one-way paratransit trip is $46.88, the MBTA is piloting a program to subsidize taxi trips for some users [5]. Despite concerns about vehicles being physically equipped and drivers having appropriate training to serve customers with disabilities, demand responsive services that allow vehicles to be shared by multiple user groups hold great promise for bringing down the cost of providing high-quality ADA paratransit service. Perhaps the changes that emerging technologies are bringing for mobility services will be a great equalizer that can afford the same transportation choices to people with disabilities as the rest of the general public. One thing is certain, the future users are going to require flexible and efficient transportation systems to meet their diverse needs.

By: Dr. Eric Gonzales

  1. American Public Transit Association (APTA) (2012). 2012 Public Transportation Factbook. Available online from: http://www.apta.com/resources/statistics/Documents/FactBook/
  2. Rahimi, M., Amirgholy, M., Gonzales, E.J. (2014). Continuum approximation modeling of ADA paratransit operations in New Jersey. Paper Number 14-4864. Transportation Research Board 93rd Annual Meeting, 12–16 January, Washington, D.C.
  3. Rahimi, M., Gonzales, E.J. (2015). Systematic evaluation of zoning strategies for demand responsive transit. Paper Number 15-4023. Transportation Research Board 94th Annual Meeting, 11–15 January, Washington, D.C.
  4. Amirgholy, M., Gonzales, E.J. (2015). Demand responsive transit systems with time dependent demand: User equilibrium, system optimum, and management strategy. Transportation Research Part B, doi:10.2016/j.trb.2015.11.006.
  5. Massachusetts Bay Transportation Authority (MBTA). Riding the T. Available online from: http://www.mbta.com/riding_the_t/accessible_services/?id=7108

Pre-signals for Transit Priority

Transit preferential treatments can reduce transit delay and therefore improve the efficiency and reliability of transit systems. Examples include dedicated bus lanes, queue jump lanes, and transit signal priority. However, these treatments are not always feasible due to lack of funding or space. In addition, they can often have detrimental impacts on other users of the system. Sustainability goals that are set by a lot of planning and transit agencies demand solutions that more efficiently utilize existing infrastructure and capacity while providing priority to transit vehicles.

Pre-signals allow for provision of priority to buses traveling on dedicated bus lanes by taking advantage of existing infrastructure and utilizing intersection capacity more efficiently. Pre-signals are additional signals placed upstream of signalized intersections to facilitate provision of some level of priority to buses, as well as other modes, by allowing them to bypass standing queues of cars. Typically, operating pre-signals require the existence of at least two lanes in the direction of travel.

However, recent work has suggested that pre-signals can aid in the temporary utilization of contra-flow lanes for transit priority provision for single lane approaches [1]. In particular, pre-signals are used upstream of the main intersection signals to allow the bus to jump the car queues and be at the front of the queue at the main signal. Pre-signals are used in combination with dedicated bus lanes when there is a need to end the bus lane in advance to allow cars to discharge from the intersection using all lanes. For example, as seen in Figure 1 the dedicated bus lane ends at some distance upstream of the intersection to allow cars to use all three lanes while discharging from the intersection.


Figure 1. Pre-signal at a three-lane approach with a dedicated bus lane.

The pre-signal works as a regular signal and is coordinated with the main signal to utilize maximum capacity. While the main signal is red, cars receive a red light at the pre-signal and are queued upstream of it. This ensures that a bus arriving during the red period can move to the stop line at the main signal and discharge immediately when the main signal turns green. Cars receive a green pre-signal such that no gaps are created in the traffic stream, and no green time at the main signal is lost when buses are not present. Regardless of the main signal’s phase, a bus approaching the intersection will trigger the pre-signal to turn red for cars, allowing the bus to move to the main signal without encountering conflicting maneuvers from cars.

An example of real-world pre-signal operations can be seen in this video. The video presents the operation of a pre-signal along Langstrasse in Zurich, Switzerland. A dedicated bus lane and one lane for cars exist upstream of the intersection but merge into a single mixed-use lane just upstream of the signalized intersection. A pre-signal at the location of the merge provides priority to buses when approaching the main signal. The pre-signal turns red when the bus is detected approaching the intersection. As a result, the bus travelling on the bus lane can bypass the queue of cars and enter the mixed-use lane at the intersection before the cars arrive. As soon as the bus bypasses the standing queue of cars, the pre-signal turns green again so that cars can proceed through the intersection after the bus.

The concept of pre-signals was first introduced to address lost time due to acceleration and perception/reaction time at the onset of green at signalized intersections and the first pre-signals were installed in Dusseldorf, Germany in 1954 [2]. This first study found that if there are only cars in a traffic stream, the equivalent of approximately 4 seconds of additional green time can be gained at intersections with the use of this type of pre-signal. More recent work has explored the use of pre-signals to increase intersection capacity by resolving various types of vehicular conflicts (e.g., between left and through moving vehicles that are either conflicting or compete for green time at the main signal) that would otherwise occur at the signalized intersection downstream [3,4]. A theoretical analysis of pre-signals for transit priority was first presented by Wu and Hounsell [5]. However, their proposed implementation included a constant pre-signal operation regardless of the arrival of a bus. To the best of our knowledge, real-world implementations of pre-signals are limited. A few locations are known in London, operating in a fashion similar to the one described in [5] and one location has been noted in Zurich, Switzerland.

We are currently working on identifying domains of application for implementation of individual transit preferential treatments or combinations of those for a variety of operating conditions for traffic and transit. Click here for a relevant presentation. 

By: Eleni Christofa, Ph.D., Assistant Professor, UMass Amherst and S. Ilgin Guler, Ph.D., Assistant Professor, The Pennsylvania State University 

[1] Guler, S.I., Gayah, V.V. and Menendez, M., 2016. Bus priority at signalized intersections with single-lane approaches: A novel pre-signal strategy. Transportation Research Part C: Emerging Technologies63, pp.51-70.

[2] Von Stein, W., 1961. Traffic flow with pre-signals and the signal funnel. Theory of Traffic Flow, Elsevier, Amsterdam.

[3] Xuan, Y., Gayah, V., Cassidy, M. and Daganzo, C., 2012. Presignal Used to Increase Bus-and Car-Carrying Capacity at Intersections: Theory and Experiment. Transportation Research Record: Journal of the Transportation Research Board, (2315), pp.191-196.

[4] Xuan, Y., Daganzo, C.F. and Cassidy, M.J., 2011. Increasing the capacity of signalized intersections with separate left turn phases. Transportation Research Part B: Methodological45(5), pp.769-781.

[5] Wu, J. and Hounsell, N., 1998. Bus priority using pre-signals. Transportation Research Part A: Policy and Practice32(8), pp.563-583.