As cars become more digital than mechanical, the inclusion of a host of smart technologies has helped accelerate the development of autonomous driving systems. However, with so many sensors and cameras on modern vehicles, there are growing concerns that such vehicles could pose a serious privacy threat as data is streamed to eliminate cloud services.
According to a review of privacy risks and mitigation strategies of smart home systems by Shahid et al. (2021), there are serious concerns about the collection and storage of data in smart devices, including those used in smart vehicles.
Why do smart vehicles need to integrate so many different sensors, do modern vehicles threaten privacy, and what can engineers do about it?
Why do smart vehicles need to integrate so many different sensors?
Intelligent vehicles rely on a range of different sensors to gather data about their environment and make decisions based on that data, with examples of sensors including cameras, radar, lidar, ultrasonic sensors and GPS. Each sensor has a specific function and is designed to collect data from a different part of its surroundings. For example, cameras are used to detect road signs, lane markings and other vehicles, while radar and lidar sensors are used to measure distances and speeds.
Combining data from different sensors is critical to generating a complete understanding of the vehicle’s environment. By combining data from various sensors, intelligent vehicles can build a detailed and accurate model of the environment in real time. This allows the vehicle to make informed decisions about its actions, such as adjusting speed, changing lanes or braking to avoid collisions.
One of the main challenges in integrating multiple sensors is the need to ensure that the data collected is accurate and reliable. This requires sophisticated electronics and methods to remove noise and interference, as well as to combine data from different sensors. For example, radar and lidar sensors can be affected by weather conditions, such as rain or fog, which can reduce their effectiveness. To address these challenges, intelligent vehicles use cutting-edge algorithms and machine learning approaches to analyze sensor data and make decisions, while using multiple sensor technologies for redundancy.
As such, the integration of multiple sensors is essential to create intelligent vehicles which are capable of autonomous driving and other intelligent functions. The use of sophisticated electronics and signal processing techniques is essential to ensure that the data collected by these sensors is accurate and reliable. As technology continues to evolve, we can expect to see even more advanced sensors and algorithms used in smart vehicles, further improving their capabilities and making driving safer and more efficient for everyone.
Are modern vehicles a threat to privacy security?
While there can be no doubting the brilliance behind modern vehicle electronics and the features they deliver, there is growing concern about the data they collect, how vehicle manufacturers use that data, and even unauthorized access by malicious parties.
In the case of dash cams, many thousands of drivers have been recorded by their vehicles without their knowledge, including Tesla CEO Elon Musk. These cameras have the ability to capture high resolution images regardless of where the vehicle is parked and in many cases store this information on a local storage device. Since such vehicles usually have some kind of cloud connectivity, a hacker only needs to find a weak spot in the connection (either directly from the vehicle or from the cloud service itself) to access video and image files. In fact, this attack was used against an unsuspecting Tesla driver who was caught completely naked while quickly grabbing items from the car.
While cameras are known to capture high-resolution images and create privacy risks, LiDAR and radar sensors may not capture images, but they can still detect objects and pose potential security risks. Zhang et al. (2020) proposed a lightweight, privacy-preserving cooperative object classification method for connected autonomous vehicles which could help address privacy and security concerns related to LiDAR and radar sensors. Using this method, connected vehicles can share information about detected objects without transmitting sensitive data, thus reducing the risk of unauthorized access to personal information.
For example, a study by researchers at the University of Michigan found that LiDAR sensors on autonomous vehicles can be used to create a vehicle “fingerprint” that can be used to track the vehicle’s movements and location (National Administration on Highway Traffic Safety, 2016). This highlights the need for advanced cybersecurity measures to protect against potential physical layer attacks such as jamming and spoofing (National Highway Traffic Safety Administration, 2016). Thus, hackers could use these sensors as motion detectors, looking for the presence of others around the vehicle (acting as a security sensor). This data could provide ideal opportunities for thieves to enter a property.
Finally, the inclusion of external and internal microphones allows modern cars to potentially record conversations (already a problem with dash cams). Although 90% of conversations are probably boring, it’s very easy to have a private conversation with someone whose content can harm either those having the conversation or others mentioned in the conversation.
What can engineers do about it?
Smart cars that integrate sensors and cameras provide various benefits such as improved safety, convenience and connectivity. However, these technologies also raise privacy and security concerns, as they have the potential to collect sensitive data and be used for surveillance purposes. Electrical engineers play a critical role in addressing these issues and ensuring that smart cars are designed with privacy as a priority.
To ensure privacy in smart cars, electronics engineers must apply design privacy principles from the very beginning of the product development cycle. This includes conducting a Privacy Impact Assessment (PIA) to identify and mitigate the privacy risks associated with the use of sensors and cameras in smart cars. They must also design systems that are secure by default, which includes using encryption, authentication, and access control to protect data from unauthorized access.
Another key consideration for electronic engineers is data minimization. Smart cars must collect and store only the data that is necessary to provide the intended functionality. Engineers must also ensure this data is anonymized or pseudonymized whenever possible to protect people’s privacy.
Electronics engineers must also prioritize transparency and user control. Smart cars must provide clear and concise notices about data collection and use and give users the ability to control their data, including providing options for users to delete their data or limit its collection.
Conclusion
In conclusion, electronics engineers play a vital role in ensuring privacy in smart cars that integrate sensors and cameras. By implementing privacy by design principles, designing systems that are secure by default, practicing data minimization, and prioritizing transparency and user control, engineers can create smart cars that are safe, comfortable, and respect individual privacy rights .