Investigation into Travel-Time Route Interpolation and Prediction, 10-R9708Printer Friendly Version
Inclusive Dates: 04/01/07 – 03/30/08
Background - Intelligent Transportation Systems (ITS) provide data to motorists about traffic conditions in the form of travel times to destinations and aberrant roadway conditions. Systems publish travel times based on sensors reporting current traffic conditions, failing to track evolving traffic conditions and possible trends found in the data. Furthermore, system functionality and accuracy degrades when traffic sensors cease to provide data, resulting in less accurate travel times or no travel times at all.
Travel times are calculated based on past traffic conditions, i.e., reported speeds of traffic as it moved past a sensor. Thus, when a motorist reaches the beginning of a route and is provided the calculated travel time, that travel time is already obsolete. The travel time was calculated in a manner oblivious to events that have occurred after the sensors reported their data.
Approach - A Travel Time Route Interpolation and Prediction (TRIP) system that takes advantage of the natural associations between data that affect travel times, the relationships between data points in the route, and their variations over time by capturing trends in the data will be better able to predict travel times in dynamic circumstances. Recent and current traffic conditions and archived sensor data and travel times can be processed to provide such traffic trends. Using this trend data, algorithms can produce travel times based on predictive and interpolative calculations. These new algorithms will be developed in staged objectives.
Accomplishments - Development of the traffic models and algorithms has been initiated. Extensive discussion with Department of Transportation traffic engineers has been held to determine the proper approach to implementing algorithms in the absence of data. Additionally, discussions about how to most effectively utilize historical data have been held which will be used in the model development.