Abstract
Vehicle telematics systems are transmitting massive volumes of data that are underutilized during the test, development and production lifecycle. Analyzing all of this data is challenging, not only due to size but also because of the variability that is encountered during on-road driving. Vehicle performance is affected by drive route, road grade, traffic, weather, vocation and driver tendencies. This study explores the efficacy of using data mining techniques to predict vehicle fuel consumption and exhaust emissions from On Board Diagnostics (OBD) data that may be harvested from telematics data stores. These data mining concepts have potential applications for eco-drive vehicle control system development, real-world emissions factor development for source apportionment models, and the analysis of Intelligent Transportation System (ITS) technology impacts on vehicle energy efficiency and emissions.