Science

Cybersecurity: A Brake for Autonomous Vehicles

GPS for navigation, biometric systems for central locking, Bluetooth, telematics for vehicle-to-vehicle communications, Light Wave Detection and Ranging (LiDAR) for obstacle detection, artificial intelligence (AI) for speed control, predictive analytics for battery replacement and several programs. systems hosted in the cloud… Autonomous vehicle technology has proven to be a gold mine in terms of data and information.

Convenience has its price

A report by the European Union Cybersecurity Agency (ENISA) and the Joint Research Center (JRC) classifies cybersecurity risks in AV as unintentional and intentional software and hardware vulnerabilities. Deliberate threats target electronic control units (ECUs), which include firmware and computer systems for various modules. ECU functions range from distance control and parking assistance to powertrain control and lane departure warning. The CAN bus protocol (controller local area network) allows the vehicle’s ECU and control modules to exchange data. If this allows the subsystems to function properly, the blocks and CAN remain vulnerable to attack.

Hackers use Bluetooth devices or USB sticks and code injection techniques to break into ECUs, CAN bus, and original equipment manufacturer (OEM) networks. For example, malicious code can be sent to an anti-theft system or a tire pressure sensor. Incorrect commands sent to the CAN bus jam the sensors, causing autonomous driving to fail or stop. Attackers can tamper with central file systems to disable the GPS system, or in the worst case, launch a ransomware attack by taking control of the antivirus and impersonating it on the OEM’s network.

In addition, suboptimal design of AI systems, inadequate training of machine learning models, and improper hardware integration can lead to unintended failures of autonomous vehicles. And the consequences of a cyberattack cannot be ignored. Given the vulnerability of smart cars, regulators are urging European manufacturers and suppliers to address cybersecurity issues.

“Safety First”

The concept of “safety by design” must be integrated into autonomous driving technology to preserve the lives and privacy of users. By integrating advanced cybersecurity measures into product design, manufacturers can mitigate deliberate attacks, artificial manipulation of AI systems, and unintended AI and machine learning vulnerabilities. Thus, this approach allows us to understand the problems of the data chain and creates an ecosystem for realizing the potential of autonomous movement. However, adequate safety testing at the design stage is still a rarity in the automotive industry. One of the reasons for this vulnerability may be the lack of own experience in the field of cybersecurity. While software development is not a major strength for automakers, deploying connected vehicles requires a team of data scientists, communication technology experts, AI developers, machine learning modellers, and analysts.

Use advanced enterprise

Collaboration with technology service providers allows OEMs to leverage cross-functional talent to build cyber-resilient AV. These companies apply end-to-end end-to-end cybersecurity strategies throughout the product lifecycle and improve the design phase to simplify reverse engineering. Digital security threat assessment and data risk analysis solutions identify, analyze and remediate vulnerabilities. Similarly, advanced access control protects order files with strong authorization methods for access and modification, while data encryption and anonymization ensure data integrity and confidentiality. In addition, attack scenario modeling tests the algorithms used to assess and mitigate risks.

Predictive analytics and simulation exercises for security risk assessment allow response teams to quickly detect abnormal vehicle behavior and misunderstandings caused by infected data or AI components, including over the air (OTA). Thus, regular security checks of built-in artificial intelligence services allow you to identify software flaws or errors. This speeds up the development of security patches for potential AI risks and emerging threats, and their rollout through an OTA update. The repository of security fixes serves as feedback for training machine learning models and updating AI systems.

As cars become smarter with built-in connectivity and artificial intelligence, cybersecurity regulations in the European Union will tighten. It is imperative that autonomous vehicles are designed not only in terms of fuel efficiency and passenger comfort, but also in terms of their safety and privacy.

Back to top button