A contemporary vehicle can be thought of as a collection of cyber-physical systems (CPS) working together to provide (i) safety and comfort to the occupants, (ii) efficient performance in terms reduced energy consumption and (iii) entertainment as well. However, this have only been possible at cost of more attack surfaces. Ample number of literature can be found where researchers have exploited these attack surfaces to launch denial-of-service , false data injection, replay attacks. We focus on developing verification methods that analyse if automotive CPSs ensure safety in the presence of an adversary.

The safety critical control softwares (for example Vehicle Stability Control , Anti-lock Braking System , Adaptive Cruise Control etc) in a modern-day car need to operate in real time. Moreover, connectivity to internet via various open ports have made these in-vehicle control systems vulnerable to both outside and inside attackers. Due to limited communication bandwidth and light weight nature of the Electronic Control Units (ECU) of a car, it is infeasible to secure every packet transmitted among the ECUs. In such cases, light weight residue-based detectors where anomalies are detected based on some pre-defined threshold, seem a promising solution. However, they suffer from false alarms. We work on designing such light weight intelligent attack detectors leveraging the concept of residue-based detectors that would ensure identification of even small attack effort as well as reduction of false alarm (for example variable threshold-based detector) in distributed automotive CPSs.