The advancement of technology in modern vehicles has brought about a significant shift in the way cars are operated, with microcontrollers utilizing the Controller Area Network (CAN) to perform crucial safety and luxury functions. However, this technological evolution has opened up vulnerabilities that could be exploited by malicious actors, leading to potential vehicle hijacking through message injection attacks since the CAN network lacks the security protocols found in drive-by-wire systems like speed control.
Various researchers have been working tirelessly to propose solutions to enhance the security features of the CAN network, including intrusion detection, encryption, and authentication. Despite these efforts, many previous works have failed to consider the practical constraints faced by automakers, thus leaving the automotive industry susceptible to cyber threats.
In response to the growing concerns surrounding the security of vehicle systems, a group of cybersecurity analysts have introduced HackCar, a revolutionary test platform designed to evaluate attacks and defenses on automotive architectures. The team behind HackCar – Dario Stabili, Filip Valgimigli, Edoardo Torrini, and Mirco Marchetti – have developed a cost-effective and fully configurable platform that simulates real-world scenarios, such as compromising an autonomous forward-collision avoidance system.
One of the major obstacles faced by investigators analyzing security risks in existing car systems has been the lack of access to actual automobile platforms, primarily due to investment barriers. HackCar addresses this issue by providing an open-sourced platform that allows researchers to replicate and expand on their work in a secure, safe, and budget-friendly environment.
The core components of HackCar include a sensing system for obstacle detection, multiple on-board controllers for sensor data analysis, actuator management, attack replication, anomaly detection, and an in-vehicle CAN network for communication among controllers using standardized data frames. This architecture enables researchers to simulate real-life scenarios, such as autonomous and manual emergency braking, while evaluating security measures through attack implementation and defensive monitoring across different layers of the vehicle system.
One of the key features of HackCar is its ability to replicate attack behaviors that can impact autonomous driving functionality. By experimenting with scenarios where the Attack Controller intercepts and overwrites critical messages, researchers have been able to validate the platform’s effectiveness in simulating realistic threats.
Overall, HackCar presents a significant advancement in the field of automotive cybersecurity, enabling researchers to prototype attacks and defenses on automotive systems in a controlled environment. By utilizing an F1-10th model and multiple automotive-grade microcontrollers, HackCar provides a comprehensive platform for studying security vulnerabilities without the need for full access to real vehicles. This breakthrough technology has the potential to revolutionize the way cybersecurity research is conducted in the automotive industry, ultimately leading to safer and more secure vehicles on the road.