Using Video Data To Measure Vehicle Operating Modes for Prediction of Emissions

Abstract

An alternative method for measuring vehicle modes of activity is examined. Emerging emissions models will require as input information on vehicle modes of operation, such as cruise, accelerations, decelerations, and idle. Measurement of vehicle modes of activity was carried out as part of a research project for which the overall objective was to determine a procedure or protocol for estimation of modal activity distributions for different road facilities. Video data collection was one of the methods used to measure modal activity. The video data collection and analysis process is described, preliminary findings and lessons learned are presented, and future needed research is identified. A major finding was that the video method enabled collection of large amounts of data that may not be possible to collect with instrumented vehicles. Although a standard for the required accuracy of modal activity data does not exist, it appears that questions regarding the accuracy of the video data may have to be resolved. Another major conclusion is that, on the basis of the video data, comparisons of modal activity distributions for different freeway sections and flow rates indicate that there are no statistically significant differences between modal activity distributions when acceleration is considered, but speed distributions in some cases varied significantly. Since the freeway sections considered do not reflect all possible freeway types and driver populations, the conclusions cannot be generalized. However, they do provide some important insights into the variability of modal activity for emissions predictions on the basis of data for commuters in California.