Intelligent Driving Pattern Recognition

Analysis of foot movements during driving has been made possible with the recent advancement of studies in the fields concerning smart driving assistance. The outcomes of previous analysis are proven to be beneficial in different areas such as in energy consumption, safety and in autonomous car feedback systems. An experiment involving embedded IMU sensors in smart shoes, combined with 4G internet connectivity and cloud storage, called Motion-core IoT has been implemented to achieve the goal in building a case library containing safe cases of drivers during driving. Approaches used in attempt to reach the goal are by conducting/implementing data preprocessing on raw data acquired during the driving experiment, clustering using K-Means algorithm, feedforward neural network and recurrent neural network, as well as suitability analysis using case-based reasoning. The goal was reached such that the system managed to successfully obtain safe cases for a test driver with an average suitability of 0.45 which is within the safe cases.