Safe and Secure Design of Connected and Autonomous Vehicles

Liu, Xiangguo

doi:https://doi.org/10.21985/n2-avgc-e934

Work

Safe and Secure Design of Connected and Autonomous Vehicles

Public

Download PDF

Download All Files (.zip)

Machine learning-based techniques have shown great promises in perception, prediction, planning, and general decision-making for improving task performance of autonomous driving. Connectivity technology has also presented great potentials in improving the safety and efficiency of transportation systems by providing information beyond the perception and prediction capabilities of individual vehicles. However, a number of challenges significantly impede their applications in realizing connected and autonomous vehicles. These challenges include (1) increasing difficulty in formally analyzing the behavior of neural network-based planners for ensuring system safety, (2) preventing over-conservative planning in dense and highly interactive traffic environments, (3) increasing complexity in analyzing system behavior and quantifying uncertainty in mixed traffic scenarios, including human-driven and autonomous vehicles, and connected and non-connected vehicles, (4) difficulty in accurately predicting surrounding vehicles' behaviors and trajectories, and (5) defending possible cyber and physical attacks on connected vehicle applications. To overcome these challenges in connected and autonomous vehicles, we propose several safety-assured planning schemes and a trust framework in this thesis. In the first work, we propose a hierarchical neural network based planner that analyzes the underlying physical scenarios of the system and learns a system-level behavior planning scheme with multiple scenario-specific motion-planning strategies. We then develop an efficient verification method that incorporates overapproximation of the system state reachable set and novel partition and union techniques for formally ensuring system safety under our physics-aware planner. With theoretical analysis, we show that considering the different physical scenarios and building a hierarchical planner based on such analysis may improve system safety and verifiability. In the second work, we propose a safety-driven interactive planning framework in mixed traffic scenarios. We identify the driving behavior of surrounding non-connected vehicles and assess their aggressiveness, incorporate the information shared by surrounding connected vehicles, and then adapt the planned trajectory for the ego vehicle accordingly in an interactive manner. The ego vehicle can proceed to execute driving tasks if a safe evasion trajectory exists even in the predicted worst case; otherwise, it can perform a less preferred behavior or follow the pre-computed evasion trajectory. Thirdly, we propose a novel speculative planning framework based on a prediction-planning interface that quantifies both the behavior-level and trajectory-level uncertainties of surrounding vehicles. Our framework leverages recent prediction algorithms that can provide one or more possible behaviors and trajectories of the surrounding vehicles with probability estimation. It adapts those predictions based on the latest system states and traffic environment, and conducts planning to maximize the expected reward of the ego vehicle by considering the probabilistic predictions of all scenarios and ensure system safety by ruling out actions that may be unsafe in worst case. For these planner designs, we demonstrate the effectiveness of our approaches and their advantages over other baselines in practical case studies of unprotected left turn, highway merging or lane changing, through extensive simulations with diverse and comprehensive experimental settings, or in real-world scenarios collected by an autonomous vehicle company. Finally, we propose an efficient dual cyber-physical blockchain framework to build trust and secure communication for CV applications. Our approach incorporates blockchain technology and physical sensing capabilities of vehicles to quickly react to attacks in a large-scale vehicular network, with low resource overhead. We explore the application of our framework to three CV applications, i.e., highway merging, intelligent intersection management, and traffic network with route choices. Simulation results demonstrate the effectiveness of our blockchain-based framework in defending against spoofing attacks, bad mouthing attacks, and Sybil and voting attacks.

Creator