In Phoenix, Arizona, it has become commonplace to pass cars equipped with spinning domes, absent a human driver. This is the work of Waymo, formerly known as the Google Self-Driving Car Project, and one of the key players in the autonomous vehicle (AV) space, providing commercial robotaxi services in Phoenix, San Francisco and Los Angeles.
Widely hailed as transportation's future, AVs employ advanced technologies like light detection and ranging, environmental controls and navigation equipment to operate without human intervention. They promise to provide mobility that “never gets drunk, tired or distracted.”
For urban areas, especially ones like Phoenix that lack comprehensive public transportation, AVs present an invaluable solution. And for those without cars or those unable to drive, AVs bridge the gap between mobility and accessibility.
In the past, automotive safety meant things like improved seatbelts and sophisticated airbag systems, but with autonomous vehicles, there’s a new layer of safety issues to consider. Simply put, the automotive industry needs to make sure AVs can navigate our roads without accidents, chaos or unexpected surprises.
History of AV and security
The dream of autonomous vehicles stretches centuries, from the time when Leonardo da Vinci sketched designs for a self-propelled cart to Norman Bel Geddes at General Motors (GM) unveiling a radio-controlled self-driving car in 1939. In the 1980s, there was a stark shift into high gear, with institutions like Carnegie Mellon University pioneering projects like Navlab and Autonomous Land Vehicle (ALV) that culminated in a cross-country autonomous drive in 1995. The new millennium brought with it a surge of interest and investment in AVs.
The U.S. Defense Advanced Research Projects Agency (DARPA) and tech behemoths like Google threw their hats into the ring, resulting in global advancements and legislative changes in support of AVs. Reflecting this unyielding growth, the AV market is poised to proliferate by $319.41 bn during 2023 to 2027, accelerating at a CAGR of 38.45%.
Real-life problems
Behind the allure of AVs lies a complex architectural system. These vehicles house extensive in-car networks, numerous sensors and intricate control systems. All these components must harmoniously collect, process and transmit data to ensure the vehicle operates safely. Despite the promise and progress, several incidents have prompted questions about AV safety and reliability:
- Uber's fatal crash: In 2018, an Uber autonomous vehicle was involved in a fatal incident in Tempe, Arizona. The driver, while found guilty of endangerment, argued that she was following Uber's monitoring instructions. The U.S. National Transportation Safety Board attributed "distraction" as the likely cause of the accident.
- Tesla's autopilot flaws: A Tesla Model Y hit a 17-year-old in North Carolina, marking one of the 736 U.S. crashes involving Tesla's Autopilot since 2019. At least 17 of these incidents were fatal.
- Google's software glitches: Google's AVs, despite their advancements, face significant technological hurdles. For instance, they find snowy conditions challenging as the snow can obscure essential visual cues. Their heavy reliance on Google Maps poses problems in undocumented areas, and they struggle in situations where traffic is managed through human hand signals.
The story surrounding vulnerability
The rise of AVs introduces a complex web of potential vulnerabilities that present grave risks. As AVs integrate sophisticated technologies, they open a multitude of attack vectors that malicious actors can exploit, threatening not only the safety of the vehicle's occupants but also that of pedestrians and other vehicles. From software glitches to hardware defects and communication breaches, each category has its unique challenges that need to be addressed.
- Software vulnerabilities: At the heart of every AV lies a complex mesh of codes and algorithms that instruct and control the vehicle on how to operate everything from navigation to collision avoidance. While these codes are designed to be error-free, they are susceptible to flaws, oversights or unintended behavior. Potential software vulnerabilities can include buffer overflows, memory leaks or even susceptibility to adversarial machine learning attacks where the vehicle's perception algorithms are deceived.
- Hardware vulnerabilities: AVs are equipped with a variety of sensors, processors and actuators. Physical tampering, malfunction or degradation of components can compromise AV decision-making or surrounding environment status. Vulnerabilities may include sensor malfunctions, wear and tear on critical parts or sabotage.
- Communication vulnerabilities: The Vehicle-to-Everything (V2X) communication paradigm is foundational to the broader integration of AVs into the transportation grid. As vehicles transmit and receive data between each other and with the infrastructure, there is a risk of interception, alteration, or malicious intent.
Technologies to achieve AV safety
The successful and safe integration of AVs into our urban fabric hinges on the development and implementation of robust safety and security technologies. Proactive threat modeling is central to this, aiming to anticipate and counter threats before they manifest. Several domains collectively work together to shape the AV safety landscape.
- Encryption: Encryption safeguards the data within AV systems to ensure integrity and confidentiality of data. Utilizing advanced cryptographic algorithms and meticulously managed encryption keys, data during transmission and at rest is safe from unauthorized access.
- Sensor technology and precision positioning: Acting as the AV's eyes and ears, sensor technologies scan the environment, detect obstacles and enable rapid decision-making to avert collisions. Complementing this, precision positioning technology locate the vehicle in space, evaluate the environment and detect nearby pedestrians.
- Intrusion detection systems (IDS): Leveraging AI and ML, IDS recognizes and can even counter an evolving threat. It can also assimilate new data and draw on historical experience to identify and mitigate potential security breaches.
- Secure boot and execution environment: This security mechanism ensures an AV boot only after authentication, offering an initial line of defense. Once in operation, the Trusted Execution Environment runs in a secure domain and isolates itself from potential threats.
- Connectivity and V2X communication: Central to AV safety, communication technologies are used to communicate with each other and the surrounding environment. They share real-time data regarding traffic conditions and hazards and foster a coordinated traffic environment to minimize vulnerabilities.
- Redundancy and fail-safe mechanisms: Serving as safety nets, fail-safe mechanisms trigger alerts when there is component failure and provide standby backup systems to sustain continuous vehicle operation.
The future of AV security technology
Currently, there is no consensus or universally accepted method for evaluating AV safety. In the future, industry leaders must think about and develop proactive technologies. Perhaps even more important, they are going to have to answer the overarching question: How safe is truly safe?
For AVs to reach their full potential, a solid framework of safety goals needs to be built. It must include fail-safe vehicle design, testing, long-term operational metrics and maintenance guidelines. As a starting point, some have proposed the following goals:
- Zero-accidents: Mitigate, if not eliminate, road accidents, setting the trajectory toward a zero-accident paradigm.
- Efficient traffic management: Seamless inter-vehicle communication to smooth traffic flow and optimize routing, reducing congestion and ensuring travel is more about the journey than the wait.
- Inclusivity and accessibility: Beyond cars, the scope of AVs extends to public transportation, offering enhanced mobility solutions for people with disabilities, ensuring a more inclusive urban transit landscape.
- Environmental stewardship: AVs stand at the intersection of efficiency and sustainability. Through optimized driving behaviors and the potential transition to zero-emission vehicles, AVs can significantly reduce carbon footprints, heralding a cleaner, greener, and safer mobility future.
Conclusion
The development of AVs, from early dreams to intricate systems, is inextricably linked to safety and security. The challenges they present create opportunities for new technologies and collaborative solutions. Manufacturers, regulators and end-users must unite to collectively outline the roadmap for an AV future that is both efficient and safe.
About the author
Emily Main holds a B.S. in Telecommunications and J.D. in Compliance Law. With extensive experience in the application of technology in business, science and finance, Main has contributed to numerous publications and conferences. She is passionate about exploring the challenges, innovations and trends related to networking and system infrastructure, and is dedicated to sharing the latest advances with professional engineering audiences.