Mining trips from telematics dataset for value-added logistics applications in asset tracking systems

With the recent improvements and cost reductions in GPS technology and wireless communication, telematics systems that integrate wireless communications with sensor-based monitoring and location aware applications have been widely deployed for mobile asset tracking and condition monitoring. In such telematics systems, mining patterns from GPS datasets and exploring the useful data for asset behaviors is vital in order to understand asset utilization, efficiency, distribution, operation, and many other important aspects in the supply chain. In this paper we mine trips based on heuristic rules from a large-scale, real world, event-based telematics dataset that is noisy and sparse in nature. A series of logistics applications built on trip analysis are presented, including ETA prediction, Mileage estimation, Backhaul detection, and Supply chain network analysis.