Scalable Urban Traffic Control

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search

Scalable URban TRAffic Control [1] (SURTRAC) is an adaptive traffic control system developed by researchers at the Robotics Institute, Carnegie Mellon University.SURTRAC dynamically optimizes the control of traffic signals to improve traffic flow for both urban grids and corridors; optimization goals include less waiting, reduced traffic congestion, shorter trips, and less pollution. The core control engine combines schedule-driven intersection control [2] with decentralized coordination mechanisms.[3][4] Since June 2012, a pilot implementation of the SURTRAC system [5] has been deployed on nine intersections in the East Liberty neighborhood of Pittsburgh, Pennsylvania.[6] SURTRAC reduced travel times more than 25% on average, and wait times were reduced an average of 40%.[7][5]A second phase of the pilot program for the Bakery Square district has been running since October 2013.[8]. In 2015, Rapid Flow Technologies[9] was formed to commercialize the SURTRAC technology. The lead inventor of this technology, Dr. Xiao-Feng Xie, states that he has no association with and does not provide technology supports for this company. [10]

Design[edit]

The SURTRAC system design has three characteristics.[5] First, decision making in SURTRAC proceeds in a decentralized manner. Decentralized control of individual intersections enables greater responsiveness to local real-time traffic conditions. Decentralization facilitates scalability by allowing the incremental addition of controlled intersections over time with little change to the existing adaptive network. It also reduces the possibility of a centralized computational bottleneck and avoids a single point of failure in the system.

A second characteristic of the SURTRAC design is an emphasis on real-time responsiveness to changing traffic conditions. SURTRAC adopts the real-time perspective of prior model-based intersection control methods [11] which attempt to compute intersection control plans that optimize actual traffic inflows. By reformulating the optimization problem as a single machine scheduling problem, the core optimization algorithm, termed a schedule-driven intersection control algorithm,[2] is able to compute optimized intersection control plans over an extended horizon on a second-by-second basis.

A third characteristic of the SURTRAC design is to manage urban (grid-like) road networks, where there are multiple competing dominant flows that shift dynamically through the day, and where specific dominant flows cannot be predetermined (as in arterial or major crossroad applications). Urban networks also often have closely spaced intersections requiring tight coordination of the intersection controllers. The combination of competing dominant flows and densely spaced intersections presents a challenge for all adaptive traffic control systems. SURTRAC determines dominant flows dynamically by continually communicating projected outflows to downstream neighbors.[3] This information gives each intersection controller a more informed basis for locally balancing competing inflows while simultaneously promoting establishment of larger "green corridors" when traffic flow circumstances warrant.

Criticism[edit]

The SURTRAC system uses closed-circuit television cameras to sense traffic conditions.[12] Surveillance of public places with CCTV networks has been criticized as enabling totalitarian forms of government by undermining people's ability to move about anonymously. Images gathered by CCTV cameras can be analyzed by automatic number plate recognition software, permitting fully automated tracking of vehicles by the license plates (number plates) they carry. Similarly, facial recognition software can analyze such images to identify and track people by the shape of their faces.

  • Clarification: In point of fact, the SURTRAC adaptive signal control system does NOT use closed-circuit television cameras to sense traffic conditions and never has. Like other adaptive signal systems on the market, SURTRAC uses a variety of commercially available traffic sensors, including video cameras, radar and induction loops. The type of video camera used does not have sufficient resolution to detect license plates or recognize faces. High-resolution CCTV camera networks are typically used by cities for Red Light running and crime prevention applications (although CCTV cameras are not currently in use in Pittsburgh or Atlanta where SURTRAC is currently deployed). In theory they could be used for traffic control detection also, but they are not generally cost-effective for this purpose. At any rate, the above claim about SURTRAC is unfounded. Please see www.rapidflowtech.com [9] for further information and find out the real facts!

It has been suggested that the benefits of traffic optimization have never been scientifically justified. It inherently favors motorized traffic over alternate modes such as pedestrians, bicyclists, and transit users and may promote more auto use.[13][14] It is suggested that an alternate approach could involve traffic calming, and a conceptual focus on the movement of people and goods rather than vehicles.

  • Clarification: It is true that many traffic optimization schemes in the past have tended to emphasize vehicles at the expense of alternative modes of travel, but there is no reason why traffic optimization cannot be multi-modal. In fact, the SURTRAC system is unique among commercially available adaptive signal control systems in that it gives explicit attention to pedestrians in circumstances where they can be detected (Other adaptive traffic signal systems ignore pedestrians completely). If the traffic sensors made available to SURTRAC are capable of distinguishing pedestrians (or for that matter bicycles, buses, and other traffic modes.), then SURTRAC can optimize with respect to priorities that are assigned to various modes (and for example give preference to pedestrians in environments where pedestrian traffic should predominate). The key challenge to multi-modal traffic optimization is multi-modal detection. For example, in the City of Pittsburgh where SURTRAC was first deployed, the only form of pedestrian detection available are pedestrian call buttons, that have to be pushed to signal presence of a pedestrian. SURTRAC takes this signaling information into account to ensure that no pedestrian waits for longer than a maximum time (typically one minute), but this is a far cry from being able to detect pedestrian volumes in different directions and integrate pedestrian flows into the optimization. Current research with SURTRAC is investigating a number of potential solutions (including smartphone-based crowdsourcing - reference to FHWA project) to the pedestrian detection problem to enable better optimization of pedestrian traffic flows. .... Traffic calming is in fact a good solution when vehicle traffic can effectively be rerouted to other roadways, but it will not solve urban congestion problems.

The SURTRAC system relies on pedestrians pushing a button in order to trigger a WALK signal, or else the pedestrian will be given a continuous "DON'T WALK" signal, despite motor traffic traveling in the same direction having a green light. Pedestrians are unlikely to ever reach an intersection that already has a walk signal unless there happened to be a pedestrian in front of them who had already hit the cross request button. This results in substantially longer wait times for pedestrians over cars to get through the same intersection, essentially making pedestrians second-class citizens of the streets. Also, many pedestrians are unaware that pushing the button is mandatory in order to receive a walk signal, and are confused when an entire light cycle occurs without ever being allowed to cross. The combination of these behaviors results in higher rates of jaywalking, leading to a less safe environment for pedestrians and higher liability for motorists.

  • Clarification: In point of fact, the above phenomenon of pushing a button in order to trigger a WALK signal is neither a feature nor a requirement of SURTRAC. It is the way that conventional traffic signals operate throughout Pittsburgh (which presumably is where this behavior has been noticed). If the City of Pittsburgh were so inclined, they could program every traffic signal to signal WALK unconditionally on every cycle, but this would come at great detriment to vehicle traffic flows.
  • There is, however, another issue, which has been a more legitimate criticism on the part of pedestrians with the way SURTRAC originally operated, that has now been corrected. Since SURTRAC is a real-time adaptive signal system it is hard to predict when a given green phase will end - it is strictly a function of current traffic flows in various directions (including pedestrians if they can be detected). When SURTRAC was originally deployed in Pittsburgh, it did not attempt to synchronize the vehicle signals with the pedestrian walk signals. So, if a pedestrian pushed the ped button, the pedestrian signal would say WALK for a predetermined amount of time and then shift to DON'T WALK (even though SURTRAC might decide to keep the signal green in the crossing direction for much longer than the required minimum). The latest version of SURTRAC fixes this problem - the system now uses the signal timing plans it generates to predict when a given green phase will end, and then uses this information to give more walk time to pedestrians. This new version of SURTRAC is operational in Atlanta GA (the most recent SURTRAC deployment as of May 2018), and the Pittsburgh deployment is currently being upgraded to this new version. This new version is expected to be operating across the entire 50-intersection Pittsburgh deployment by July 2018.

Because they require a continuous supply of electricity, automatic traffic signals are not suitable for use in places where the electric supply is sporadic or nonexistent. For example, traffic in Pyongyang, North Korea is guided by government workers who stand in the intersections under umbrellas.[15][16]

When drivers become accustomed to automated traffic signals, they may forget how to properly yield the right of way, so that when the electric supply is interrupted, as when a disaster occurs, traffic may not flow as well as if the signals had never been used.[17][original research?] This effect could conceivably delay evacuation or impede the movement of emergency vehicles.

The introduction of traffic signals and associated laws may undermine democracy by conditioning citizens to reflexively obey the signal lights.[18][19][original research?]

Roundabouts are an alternative to signaling systems. At a roundabout, motor traffic may not have to come to a stop (so that drivers' time and fuel may be saved), and crossing for pedestrians may be easier. Studies of intersections converted to roundabouts have found reductions in the frequency and severity of accidents.[20] However, these benefits may not be realized if a roundabout is poorly designed. Roundabouts typically require a greater land area than intersections, so doing such conversions in heavily built-up areas may imply demolition of adjacent structures.

See also[edit]

Other adaptive traffic control systems[edit]

References[edit]

  1. ^ Xiao-Feng Xie, S. Smith, G. Barlow. Smart and Scalable Urban Signal Networks: Methods and Systems for Adaptive Traffic Signal Control. U.S. Patent No. 9,159,229, 2015.
  2. ^ a b Xiao-Feng Xie, Stephen F. Smith, Liang Lu, Gregory J. Barlow. Schedule-driven intersection control. Transportation Research Part C: Emerging Technologies, 2012, 24: 168-189.
  3. ^ a b Xiao-Feng Xie, Stephen F. Smith, Gregory J. Barlow. Schedule-driven coordination for real-time traffic network control. International Conference on Automated Planning and Scheduling (ICAPS), Sao Paulo, Brazil, 2012: 323-331.
  4. ^ Hu, H-C and S.F. Smith, “Softpressure: A Schedule-Driven Backpressure Algorithm for Coping with Network Congestion”, Proceedings 27thInternational Joint Conference on Artificial Intelligence, Melbourne, Australia, August 2017
  5. ^ a b c Stephen F. Smith, Gregory J. Barlow, Xiao-Feng Xie, Zachary B. Rubinstein. Smart urban signal networks: Initial application of the SURTRAC adaptive traffic signal control system. International Conference on Automated Planning and Scheduling (ICAPS). Rome, Italy, 2013.
  6. ^ Stephen F. Smith, Gregory Barlow, Xiao-Feng Xie, and Zack Rubinstein. SURTRAC: Scalable Urban Traffic Control. Transportation Research Board 92nd Annual Meeting Compendium of Papers, 2013.
  7. ^ Walters, Ken (October 16, 2012). "Pilot Study on Traffic Lights Reduces Pollution, Traffic Clogs". CMU website. Carnegie Mellon University. Retrieved January 31, 2013.
  8. ^ Barlow, G.J., S.F. Smith, X-F Xie and Z.B. Rubinstein, “Real-Time Traffic Control for Urban Environments: Expanding the Surtrac Testbed Network”, 2014 World Congress on Intelligent Transportation Systems, Detroit, MI, September 2014.
  9. ^ a b "Rapid Flow Technologies". www.rapidflowtech.com. Retrieved 2018-06-02.
  10. ^ Xiao-Feng Xie. "Statement on the Scalable Urban Traffic Control Technology". Retrieved 2018-07-03.
  11. ^ M. Papageorgiou, C. Diakaki, V. Dinopoulou, A. Kotsialos,and Y. Wang. Review of road traffic control strategies. Proceedings of the IEEE, 2003, 91(12):2043–2067.
  12. ^ Walters, Ken (2012-10-16). "Smart Signals: Pilot Study on Traffic Lights Reduces Pollution, Traffic Clogs". CMU Piper. Retrieved 2013-01-28.
  13. ^ Michael J. Vandeman, "Is Traffic Signal Synchronization Justifiable?", April 15, 1994
  14. ^ Meyer, Robinson. "Sorry, Los Angeles: Synchronizing Traffic Lights May Not Reduce Emissions". Theatlantic.com. Retrieved 28 January 2013.
  15. ^ "Traffic Control Platform beneath Umbrella Installed at Intersections of Pyongyang". KCNA. 2009-08-13. Retrieved 2013-01-29.
  16. ^ "Meet The Ladies Of The Pyongyang Traffic Bureau". Jalopnik.com. 2010-03-05. Retrieved 2013-01-29.
  17. ^ Remaly, Jake (2012-10-29). "Fallen Trees, Wires, Poles Wreak Havoc on I-287, Local Roads in Morris County - Jefferson, NJ Patch". Jefferson.patch.com. Retrieved 2013-01-29.
  18. ^ "Houghton Mifflin Textbook - Chapter Outline". College.cengage.com. Retrieved 2013-01-29.
  19. ^ "U. Houston : PSYCH 1300 : Chptr_07". Coursehero.com. Retrieved 2013-01-29.
  20. ^ "Q&A: Roundabouts". Iihs.org. Retrieved 2013-01-29.

External links[edit]