Design of an Intelligent Vehicle Accident Detection System

The number of cars on the roads increases annually, and so does the number of vehicle accidents. It often happens that in the event of an accident, it is only at the accident scene that the emergency services find out the actual state of the accident. Many times, before arriving at the accident site, they do not even know how many people were in the vehicle at the time of the accident, what the speed of impact was, how many times and whether the vehicle overturned at all and other essential parameters. In this paper, we propose a solution for accident detection by using a Raspberry Pi 4B+ microcomputer with PiCAN 3 board, which is connected to the vehicle Controller Area Network (CAN bus) via the OBD-II port. An Arduino microcontroller board is included to operate an accelerometer to measure the acceleration and inclination of the vehicle. Data collected from the vehicle are transmitted to a mobile application via Bluetooth low energy (BLE) technology, which collects and evaluates the data about a possible accident. This data is sent to the emergency center server, stored in the MongoDB database and the severity of the accident is evaluated. Finally, according to the received data, an appropriate number of emergency vehicles might be dispatched to the scene of the accident.