This project aims to achieve security and resilience for future vehicular systems. It tackles the limitations of current approaches to cybersecurity in the transport sector by introducing new modelling and verification techniques based on digital twins. The approach contains multiple design perspectives, each one encompassing a property of concern.
The transport sector is expanding its capabilities to allow autonomy, cooperation and reliability. Advancements in sensing technology and the availability of computation and communication devices on vehicles enables progress. The widespread adoption of these systems is dependent on them being reliable, safe, secure, and resilient. However, current engineering practice cannot guarantee that these properties are met at all times. The consequence is that issues are not realised during the design phase of these systems. Problems later emerge after deployment, often with tragic consequences.
This project evaluates the adequacy of a new approach by working on two use cases provided by our industrial partners: a connected car scenario (NXP) and a fractionated aerial system (Blue Bear). In the first use case, the project develops techniques to guarantee the resilience of on-board AI against cyberattacks. In the second use case, the concern is detecting and preventing cognitive dissonance when a human operator cooperates with a large swarm of highly autonomous drones to complete a mission.