No abstract

Automated Lane Centering (ALC) systems are convenient and widely deployed today, but also highly security and safety critical. In this work, we are the first to systematically study the security of state-of-the-art deep learning based ALC systems in their designed operational domains under physical-world adversarial attacks. We formulate the problem with a safety-critical attack goal, and a novel and domain-specific attack vector: dirty road patches. To systematically generate the attack, we adopt an optimization-based approach and overcome domain-specific design challenges such as camera frame inter-dependencies due to dynamic vehicle actuation, and the lack of objective function design for lane detection models.We evaluate our attack method on a production ALC system using 80 attack scenarios from real-world driving traces. The results show that our attack is highly effective with over 92% success rates and less than 0.95 sec average success time, which is substantially lower than the average driver reaction time. Such high attack effectiveness is also found (1) robust to motion model inaccuracies, different lane detection model designs, and physical-world factors, and (2) stealthy from the driver's view. To concretely understand the end-to-end safety consequences, we further evaluate on concrete real-world attack scenarios using a production-grade simulator, and find that our attack can successfully cause the victim to hit the highway concrete barrier or a truck in the opposite direction with 98% and 100% success rates. We also discuss defense directions.

Machine learning techniques, particularly those based on deep neural networks (DNNs), are widely adopted in the development of advanced driver-assistance systems (ADAS) and autonomous vehicles. While providing significant improvement over traditional methods in average performance, the usage of DNNs also presents great challenges to system safety, especially given the uncertainty of the surrounding environment, the disturbance to system operations, and the current lack of methodologies for predicting DNN behavior. In particular, adversarial attacks to the sensing input may cause errors in systems' perception of the environment and lead to system failure. However, existing works mainly focus on analyzing the impact of such attacks on the sensing and perception results and designing mitigation strategies accordingly. We argue that as system safety is ultimately determined by the actions it takes, it is essential to take an end-to-end approach and address adversarial attacks with the consideration of the entire ADAS or autonomous driving pipeline, from sensing and perception to planing, navigation and control. In this paper, we present our initial findings in quantitatively analyzing the impact of a type of adversarial attack (that leverages road patch) on system planning and control, and discuss some of the possible directions to systematically address such attack with an end-to-end view.

No abstract

No abstract