The Privacy Dilemma: How Autonomous Cars May Impact Your Personal Data

As cars become more digital than mechanical, the inclusion of numerous smart technologies has helped to accelerate the development of autonomous driving systems. However, with so many sensors and cameras on modern vehicles, there are increasing concerns that such vehicles may introduce a serious threat to privacy as data is streamed to remove cloud services. 

According to a review of privacy risks and mitigation strategies for smart home systems by Shahid et al. (2021), there are significant concerns regarding data collection and storage 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 this?

Why do smart vehicles need to integrate so many different sensors?

Smart vehicles rely on a range of different sensors to gather data about their environment and make decisions based on this 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 the vehicle's 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 various sensors is crucial for generating a complete understanding of the vehicle's environment.  By combining data from different sensors, smart vehicles can build a detailed and accurate model of their environment in real-time. This allows the vehicle to make informed decisions about its actions, such as adjusting its 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 needs advanced electronics and methods to remove noise and interference, as well as to merge data from various 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, smart vehicles employ cutting-edge algorithms and machine learning approaches to analyse sensor data and make decisions while also using multiple sensing technologies for redundancy.

As such, the integration of multiple sensors is essential for creating smart vehicles that are capable of autonomous driving and other intelligent features. 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 being used in smart vehicles, further enhancing their capabilities and making driving safer and more efficient for everyone.

Do modern vehicles pose a security threat to privacy?

While there can be no doubt of the brilliance behind modern electronics in vehicles, and the features they present, there is an increasing concern regarding the data they collect, how vehicle manufacturers use that data, and even unauthorised access by malicious parties.

In the case of vehicle cameras, 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-definition images, regardless of where the vehicle is parked, and in many cases, store this information on a local storage device. As such vehicles usually have some kind of cloud connectivity, a hacker only needs to find a weak spot in the connection (either from the vehicle directly or the cloud service itself), to access video and image files. In fact, this attack was used on one unsuspecting Tesla driver who was caught fully nude while quickly fetching items from the car.

While cameras are known to capture high-definition images and pose privacy risks, LiDAR and radar sensors may not capture images but 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 that could help address privacy and security concerns associated with LiDAR and radar sensors. By using this method, connected vehicles can share information about detected objects without transmitting sensitive data, thereby reducing the risk of unauthorised access to personal information. 

For example, a study by researchers at the University of Michigan found that LiDAR sensors on autonomous vehicles could be used to create a "fingerprint" of the vehicle, which could be used to track the vehicle's movements and location (National Highway Traffic Safety Administration, 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 would be able to utilise 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 burglars to enter a property. 

Finally, the inclusion of microphones both externally and internally allows modern cars to potentially record conversations (this is already an issue with dashcams). While 90% of conversations are arguably boring, it is very easy to have a personal conversation with someone whose contents could be damaging either to those having the conversation or others mentioned in the conversation.

What can engineers do about this?

Smart cars that integrate sensors and cameras provide various benefits, such as improved safety, convenience, and connectivity. However, these technologies also raise concerns about privacy and security as they have the potential to collect sensitive data and be used for surveillance purposes.  Electrical engineers have a crucial role in tackling these issues and making sure smart cars are designed with privacy as a priority.

To ensure privacy in smart cars, electronic engineers should implement privacy-by-design principles from the outset of the product development cycle. This involves conducting a privacy impact assessment (PIA) to identify and mitigate privacy risks associated with the use of sensors and cameras in smart cars. They should also design systems that are secure by default, which includes using encryption, authentication, and access controls to protect data from unauthorised access.

Another key consideration for electronic engineers is data minimisation. Smart cars should only collect and store the data that is necessary to provide the intended functionality. Engineers should also ensure that data is anonymized or pseudonymized when possible to protect the privacy of individuals.

Electronic engineers should also prioritise transparency and user control. Smart cars should provide clear and concise notices about data collection and use and give users the ability to control their data, and this includes providing options for users to delete their data or limit its collection.

Conclusion

In conclusion, electronic 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, practising data minimisation, and prioritising transparency and user control, engineers can create smart cars that are safe, convenient, and respectful of individual privacy rights.