Online Verification of Automated Road Vehicles Using Reachability Analysis

Abstract

An approach for formally verifying the safety of automated vehicles is proposed. Due to the uniqueness of each traffic situation, we verify safety online, i.e., during the operation of the vehicle. The verification is performed by predicting the set of all possible occupancies of the automated vehicle and other traffic participants on the road. In order to capture all possible future scenarios, we apply reachability analysis to consider all possible behaviors of mathematical models considering uncertain inputs (e.g. sensor noise, disturbances) and partially unknown initial states. Safety is guaranteed with respect to the modeled uncertainties and behaviors if the occupancy of the automated vehicle does not intersect that of other traffic participants for all times. The applicability of the approach is demonstrated by test drives with an automated vehicle of the Robotics Institute at Carnegie Mellon University.

A Formally Verified Checker of the Safe Distance Traffic Rules for Autonomous Vehicles

“…We assume that the verification could be more organised by following a dedicated calculus for hybrid systems [20], which could be embedded in Isabelle/HOL. Our checker could also be extended with reachability analysis [2,13] in order to verify a continuous trace. Lastly, aligned with our previous work in formalisation of traffic rules [23], we wish to increase the number of formalised traffic rules such that the liability issue can be deduced automatically with our checkers.…”

Formalising Traffic Rules for Accountability of Autonomous Vehicles

“…As for rule Φ 5 , the main difficulty lies in the notion of "safety reasons" which is very abstract. Nevertheless, safety reasons can potentially be concretised by prediction under uncertainty using reachability analysis [12].…”

“…Formal verification for autonomous vehicles can be broadly classified into two categories: set-based and logic-based approaches. With set-based approaches, such as in [12], a requirement is formalised by specifying a set of acceptable behaviours. Meanwhile, with logic-based approaches, such as [13]- [15], a requirement is formalised by translating it into a logical sentence directly.…”

A Model-Based Approach to Dynamic Self-assessment for Automated Performance and Safety Awareness of Cyber-Physical Systems

“…A great many techniques have been developed and explored for advanced verification and validation of complex systems, both statically at design-time and dynamically at run-time. One particularly promising type of model checking is the so-called on-line model checking, as explored in [25]. In this approach, the resulting combined state space is continuously monitored against critical safety properties, or used to compute safe trajectories.…”