Analysis, characterization, and visualization of freeway traffic data in Los Angeles

Presented is an analysis of a large volume of readily available loop detector based traffic data for the Los Angeles and Ventura Counties. The data suggests that the daily temporal variation of congestion along any directional road segment can be characterized quite well by a 10-parameter function. The function is shown to be suitable for use in the classification of road segments, such as having morning but no afternoon or evening congestion, as well as for the purpose of improving real-time forecasts of congestion ahead for use in generating dynamic real-time minimum estimated time-of-arrival turn-byturn navigation instructions. Automation of the process allows for the characterization of all of the Los Angeles and Ventura Counties and can be applied to any metropolitan area having similar data. Several interactive and dynamic visualization tools using Google Earth are also developed and presented.