How Embedded Tech Is Securing Connectivity in Autonomous Vehicles
February 26, 2021
Today’s vehicles are progressively more advanced than ones from past generations. They can help a driver steer into a tight parking place, warn them of an impending collision or alert them that they’ve veered into the next lane.
As these automobiles become more high-tech, the likelihood goes up of cybercriminals successfully hacking them. Security researchers have demonstrated how hackers could remotely stop a moving car, for example.
It’s easy to envision how the risks go up with autonomous vehicles. They have gigantic assortments of sensors, cameras, and other technology to help them navigate crowded streets without a person’s hand ever touching the steering wheel. If a hacker gains control of an autonomous car, a catastrophe could happen before a human could react to it or even know what happened.
Fortunately, embedded tech could make these vehicles safer. Here’s a look at how that’s happening and what’s in store for the future.
Designing an Engineering Framework
A new European Union-funded initiative called XANDAR (X-by-Construction Design Framework for Engineering Autonomous & Distributed Real-Time Embedded Software Systems) aims to refine the steps taken when designing software for embedded systems. The researchers are looking beyond autonomous vehicles, but they clarified that this effort applies to them.
Jürgen Becker is coordinating the project. He explained the significance of the work by saying, “We will make available to system engineers and programmers a standardized toolchain to control automatized hardware/software integration according to all relevant criteria at early stages of the design process.”
He also mentioned that this system considers both functional and nonfunctional requirements of an embedded system. Besides making driverless cars safer, this approach could lower overall costs.
If engineers have a clear framework to follow during all stages of product development, it’s easier for teams to feel confident that they have not overlooked any crucial steps. It also saves time by reducing the chances of needing to go back and tweak things after testing identifies harmful weaknesses with a car’s functionality.
Reducing the Data Collected During Operations
Besides protecting the cars themselves from malicious hacking, the people who design and develop autonomous vehicles must minimize the data collected by the associated embedded systems. That approach protects user privacy and prevents too much information from falling into the wrong hands. For example, vehicle manufacturers and mapping specialists spend time and money developing detailed maps. Keeping the mapping data well-protected is a business interest.
The European Data Protection Board urges a data protection by design approach for the embedded systems used in connected cars. It also recommends that designers create default settings to protect privacy and offer more transparency about how data is collected and used.
Security is not the only challenge associated with 3D mapping. For example, the associated system must tolerate harsh environments, function with minimal power requirements, offer a high-performance computing platform and fit within the limited space of a vehicle. However, taking a foundational approach to security will help build a superior car that’s less likely to show unexpected design or engineering flaws.
Following the Applicable Safety Best Practices
Adhering to security safeguards spans beyond autonomous cars. For example, IT professionals follow standards to monitor and reduce their risks of cybersecurity attacks. Doing that minimizes issues associated with malicious or unintentional data breaches or poorly secured systems. Automobile engineers could take a similar approach to keep their embedded systems secure.
While speaking at the 2019 Embedded Safety and Security Summit, senior safety and security engineer David Johnson explained a two-prong approach to ensuring that car safety systems function as intended. First, engineers must make them consistently reliable through things like smart design, testing and verification. Next, they must adopt a cybersecurity approach, ensuring that the system does not have unaddressed vulnerabilities that help hackers get access.
This suggested approach highlights how keeping autonomous cars does not necessarily mean using the most advanced systems available. Rather, it requires adhering to sound engineering and design practices throughout a system’s development.
Using Artificial Intelligence to Bolster Car Safety
A pervasive argument about autonomous cars is that they could reduce accidents caused by human errors. An artificial intelligence (AI) system won’t become distracted by a fussy child in the backseat or swerve briefly into the adjacent lane while fiddling with the audio system’s dial. However, some researchers worry about AI making ethical decisions. When faced with life-or-death circumstances, would they make similar judgments to humans?
More research must occur in that area before people draw conclusions. However, AI may complement security efforts. A company called Tiempo Secure recently received recognition during a cybersecurity challenge hosted by the French government. The company’s technology creates more resilient embedded systems by automatically connecting them to an AI server. That setup allows better detection of cyberattacks, including emerging threats. It also allows authorized parties to regain control of a compromised product.
Taking a Comprehensive Approach Meets Goals
This overview emphasizes that there is no single best way to use embedded technology to secure autonomous car systems. However, using a methodical approach to assess weaknesses, examine the type of data collected and follow best practices during design and engineering will improve the chances that embedded technology supports security efforts rather than compromising them.