Design and implementation of a novel algorithm to smart tachograph for detection and recognition of driving behaviour
Abstract
Losses in accidents involving heavy vehicles, where the use of tachograph devices is mandatory, are higher than other vehicles in terms of death, injury and cost. Existing tachographs do not provide any information on how to make turns and lane changes. Aggressive driver behaviour can cause a significant portion of traffic accidents and fuel consumption. In this study, a low-cost inertial measurement unit (IMU) sensor module is mounted to the tachograph to detect lateral manoeuvres. To detect manoeuvres such as right-left turns and lane changes with high accuracy, the edges of the events are firstly captured with Gyroscope-Z data and start/end of the manoeuvres are detected. Then, a new algorithm is proposed that scores the turn manoeuvres by combining the averages of the Gyroscope-Z, Accelerometer-X and speed data in the manoeuvre. In practice, it has been observed that the algorithm approaches 100% accuracy in detecting turns and 88% accuracy in lane changes.