Special Session

22.3 - Runtime Monitoring for Safety of Intelligent Vehicles

1:30 PM–3:00 PM Jun 26, 2018 (America - Los Angeles)

Room 3020

Advanced Driver-Assistance Systems (ADAS), autonomous driving, and connectivity have made automotive design more complicated than ever. Formal verification can be applied to guarantee the functional correctness, but the scalability is limited due to the high complexity. To be applicable with high complexity and limited resource in practice, one alternative is to apply runtime monitoring which detects unsafe states (requirement violations) of systems. In this work, we first investigate appropriate formal property specification languages, such as Linear Temporal Logic (LTL) and Metric Temporal Logic (MTL), for automotive systems. We then explore existing tools which can automatically or semi-automatically generate codes of runtime monitors. With languages and tools, we perform a case study showing how runtime monitoring can assist the integration of an erroneous-start prevention system and a pile-up mitigation System. This is an initial step to utilize runtime monitoring to solidify the design safety of intelligent vehicles.