Bottleneck Identification and Forecasting in Traveler Information Systems

Major urban areas experience active bottlenecks on a regular basis. Recently, the US Federal Highway Administration launched a new initiative aimed at identifying key bottlenecks in each state. In Oregon, a regional archive of freeway data from inductive loop detectors (PORTAL) is currently leveraging bottleneck identification and prioritization efforts. In addition to products that are attractive to transportation operators, traveler information systems may benefit from identification, historical analysis, and prediction of bottleneck conditions. This paper discuses our efforts toward expanding traveler information tools already available in PORTAL, in particular, historical analysis and display of active bottleneck features, activation detection in live data, and forecasting of relevant features such as shockwave propagation and estimated onset time upstream. By providing automatic mechanisms that incorporate learned features from historical data into live displays, traveler information systems may better serve users as they are told when and where to expect recurrence in bottlenecks, in addition to detecting non-traditional bottlenecks, such as those produced by incidents or atypical seasonal congestion.