Modeling IoT Based Automotive Collision Detection System Using Support Vector Machine

Due to the rise in automotive accidents across the globe, a cheap and reliable system is needed that can be retrofitted in any type of vehicle and can monitor road collision events. This research is aimed to develop an IoT system that uses contemporary smartphone’s intrinsic sensors to accurately report vehicle collision accidents on the road. Absolute linear acceleration (ALA) and the speed of the vehicle have been used to train and test our Support Vector Machine (SVM) based collision detection model. During the testing, the accuracy of the model was found to be very high with a MAPE (mean absolute percentage error) of 0.6%.