Runtime Monitoring and Fault Detection for Neural Network-Controlled Systems⁎⁎This work was supported in part by the Leverhulme Trust Early Career Fellowship under Award ECF-2021-517, and in part by the UK Royal Society International Exchanges Programme under Award IES\R3\223168. Xianxian Zhao was supported by the SEAI (Sustainable Energy Authority of Ireland) under RD&D Award 22\RDD\776.

There is an emerging trend in applying deep learning methods to control complex nonlinear systems. This paper considers enhancing the runtime safety of nonlinear systems controlled by neural networks in the presence of disturbance and measurement noise. A robustly stable interval observer is designed to generate sound and precise lower and upper bounds for the neural network, nonlinear function, and system state. The obtained interval is utilised to monitor the real-time system safety and detect faults in the system’s outputs or actuators. An adaptive cruise control vehicular system is simulated to demonstrate effectiveness of the proposed design.