Risk Assessment in the Context of Dynamic Reconfiguration of Level of Driving Automation in Highly Automated Vehicles

Advanced Driver Assistance Systems (ADAS) constitute a field that continues to attract immense research BY promising significant advantages and novelties to the manner in which we drive vehicles, facilitating several of the driver's operations and the passengers' journey, as well as protecting the vehicle from undesired situations. With the advent of Automated Vehicles (AVs), the research and development in ADAS will be intensified, so as to holistically undertake the responsibility of getting a vehicle safely from one point to another point. Risk assessment and reliability analysis are a cornerstone of the evaluation of ADAS and their probability of success in completing a prescribed mission of an AV. In this paper, the classical Failure Mode and Effect Analysis (FMEA) technique is applied to investigate the risk assessment regarding real time adaptation of the Level of Driving Automation (LoDA) in AVs. This analysis is crucial as high risky events are evolved in the transition mode of LoDA related to hardware, sensors, software failures, front obstacles or crashes, dense traffic congestion, adverse weather or road conditions, etc. Through this analysis, an efficient approach is developed by exploring the reliability of the LoDA transition in an AV operation and its impact on the behaviour of both the AV and the driver.