Intelligent Vehicle Monitoring for Safety and Security
To be published on April 2, 2019 by SAE International in United States
The challenges posed by connected and autonomous vehicles fall beyond the scope of current version of ISO 26262. According to the current functional safety standard, human intervention defines the degree of severity of the fault. Since the driver involvement in CAVs will decrease in future, this classifies all malfunctions/faults as ASIL D. On the other hand, CAVs bring additional capabilities such as advance sensors, telematics-based connectivity etc. which can be used to devise efficient approaches to address functional safety challenges. The caveat to these additional capabilities are issues like cybersecurity, complexity, etc. This paper will present a systematic approach to understand challenges and propose a potential solution to handle faults/malfunctions in CAVs. This approach requires a framework, presented in this paper, to deal with the functional safety challenges when the driver is not in the loop. The framework introduces the concept of ‘Smart Diagnostics’ algorithms that utilize the additional set of sensors and connectivity available in CAVs. These smart algorithms may be model-based for the on-board systems (e.g. motors, sensors etc.) or machine learning based algorithms can be used to diagnose the processes` malfunction (E.g. lane changing, overtaking, following the speed limits, parking, etc.) in CAVs. The framework will build upon the lessons learned from current L2 technology out on the road. On top of it, smart diagnostics needs to be created with a level of cybersecurity awareness.