Autonomous systems require computational resources to actively make decisions on vehicle operations without human input. These systems rely on sensing technologies, physics, and decision engines. While operating heavy vehicles autonomously is a feat and shows immense promise, the current implementation and deployment of these autonomous systems have enjoyed a non-adversarial environment. However, to grow and scale the AMT program, the system will need to be robust in an adversarial environment. The cybersecurity of the system needs to be assessed and the threats and potential attack vectors need to be explored and any discovered vulnerabilities will need to be addressed.

A potential effect of a cybersecurity attack on the AMT barrier would lead to a traffic crash. Interestingly, the AMT barrier is designed to dissipate the energy of a vehicle that potentially strikes it. In fact, this scenario may be an inevitability, in which the barrier technology should mitigate the potential tragic abrupt dissipation of energy. However, when crashes happen, the sequence of determining crash causation starts with a traffic crash reconstruction. This process gathers physical and digital evidence associated with the crash and reconstructs the events that lead to the crash. This process is not well understood with AMT since the technology is new and an actual traffic crash has not happened yet.

The connection between cybersecurity and crash reconstruction is two-fold: 1) much of the evidence needed to accurately reconstruct the traffic crash is digital in nature and needs to be preserved in a secure manner, and 2) a potential of a cybersecurity breach could lead to a crash. Therefore, it is critically important to undergo a study where a forensic investigation is conducted (i.e. an exercise in data gathering from a simulated crash event).

It is important to understand how to investigate a crash when it happens. What do crash investigators do when they first arrive on scene and gather data? The data they need are both physical, (i.e. photographs and measurements), and electronic from event data recorders, video feeds, and operating logs. How does law enforcement request the appropriate data and how accurate is it?