Shailja Thakur: “Security and interpretability in automotive systems”

Shailja Thakur will present her work on November 10th (2021) at 3:00PM.

It will be online at https://ecolepolytechnique.zoom.us/j/3910379384?pwd=MDYrT0ZJOGtRbnllMDViRUU4em1JZz09

 

Title

Security and interpretability in automotive systems

Abstract

CAN is the most commonly found bus protocol in automotive systems. The two-wire bus protocol helps accomplish sophisticated vehicle services in real-time through complex interactions between hardware components. However, the lack of any sender authentication mechanism in place makes CAN susceptible to security vulnerabilities and threats. To address the insecure nature of the system, one of my research contributions is a sender-authentication technique that uses power consumption measurements of the ECUs and a classification model to determine the transmitting states of the ECUs. The method’s evaluation in real-world settings shows that the technique applies in a broad range of operating conditions and achieves good accuracy.
A key challenge of machine-learning-based security controls is the potential of false positives. A false-positive alert may induce panic in operators, lead to incorrect reactions, and in the long run cause alarm fatigue. For reliable decision-making in such a circumstance, knowing the cause for unusual model behaviour is essential. But, the black-box nature of these models makes them uninterpretable. Therefore, another contribution of my research explores explanation techniques for inputs of type image and time-series that (1) assign weights to individual inputs based on their sensitivity toward the target class, (2) and quantify the variations in the explanation by reconstructing the sensitive regions of the inputs using a generative model.
In summary, my research work is on the security and interpretability in automotive systems, which can also be applied in other settings where safe, transparent, and reliable decision-making is crucial.

Bio

Shailja Thakur is a Ph.D. candidate at the University of Waterloo, advised by Prof. Sebastian Fischmeister. Before joining the Ph.D. program, she worked for Compuware on driving behaviour analytics for GM and Ford. She received her master’s from IIIT Delhi, India, where she worked with Dr. Amarjeet Singh on apportioning home energy consumption at individual and appliance levels in home and dorm settings. Before this, she received her bachelor’s in CS from the University School of Information Technology, GGSIPU, New Delhi.

Comments are closed.