ELSTS, A., MEDNIS, A., SELAVO, L. 2014. Bayesian Network Approach to Vehicle Mode Monitoring Using Embedded System with 3-axis Accelerometer. International Journal of Imaging and Robotics, 12(1), pp. 67-80.
Vehicle activity monitoring is an important problem in the transport domain: the activities have impact on the fuel consumption of the vehicle, and therefore its economical feasibility. Activity detection usually is performed using relatively expensive and complicated GNSS receivers. We propose a Bayesian network based algorithm for vehicle mode detection. The algorithm can reliably differentiate between standing and moving modes in a diverse set of test scenarios. It is suitable for a low-cost hardware consisting of low-power microcontroller and uses 3-axis accelerometer as the sole sensor.