Real-Time Prediction and Decision Making in Connected and Automated Vehicles Under Cyber-Security and Safety Uncertainties

Our current transportation system is on the brink of transforming into a highly connected,automated, and intelligent system as a result of the rapid emergence of connected andautomated vehicles (CAVs). CAVs, with various levels of automation, are expected toincrease overall road safety, reduce travel time, improve comfort, improve fuel efficiency, anddecrease fatal accidents in the near future. CAVs use a combination of cameras, ultrasonicsensors, and radar to build a digital map of their surroundings and operate the vehicleaccordingly. As a result, there are numerous sources of information that can be manipulated,with malicious or non-malicious intent, which may result in dangerous situations. Althoughthe ever-increasing use of CAV technologies in vehicles are expected to have numerousadvantages, they can give rise to new challenges in terms of safety, security, and privacy.As evident by recent crash records and experiments successfully conducting cyber attacks onvehicles, the currently available autonomous systems lack the ability to fully handle novel,complex situations. Hence, the potential drawbacks of CAVs are not negligible and shouldnot be ignored. In this study, we investigate the real-time prediction and decision makingin CAVs under cyber-security and safety uncertainties.