A smart vehicle refers to a computer-controlled vehicle that can guide itself, familiarize itself with its surroundings, make decisions, and fully operate without any human interaction. Smart vehicles are most sought after for both driver and vehicle safety, with the prominent reason being to minimize road accidents caused due to drivers’ negligence like using mobile phones while driving, over-speeding, and other distractions. Connected vehicles have a lower chance of making mistakes, due to autonomous vehicle safety, which helps lessen life-threatening circumstances on the road. The autonomous and connected cars concept provides better utilization of resources and a hassle-free driving experience. The use of autonomous vehicles can help in coordinating traffic on highways and reduce tailbacks. Smart parking, incident warning, emergency braking, and semi-automatic and fully automatic (limited) pilot-driving are among the features offered by the leading car vendors, all of which have increased competition in the connected car industry.
Cars that are in the market now come with a keyless operation and are designed to prevent car theft. The car can identify its legitimate owner using biometric features. Smart vehicles host a myriad of sensors and actuators and share a great deal of data. As a result, the amount, speed, quality, heterogeneity, and real-time nature of data must be taken into consideration when designing an autonomous vehicle. Smart vehicles provide sophisticated features for the car and its owner, as part of cybersecurity in autonomous vehicles. However, certain reports and cyber experiments that were conducted on smart vehicles for data security and privacy concerns are not to the liking of autonomous vehicle enthusiasts.
In 2019, according to reports on automotive cyber-attacks, a hacker who called himself L&M gained access to two GPS monitoring applications. This allowed him to track the real-time whereabouts of thousands of cars across different nations and even cut off their engines while on the road. According to an internet post, the hacker said that he was able to follow automobiles in numerous nations across the world using the automobiles’ smart apps, including India, Morocco, the Philippines, and South Africa. He exploited a basic vulnerability in the GPS tracking of the vehicles. By brute-forcing millions of usernames with the default password, he was even able to access the personal (PII) details. Around November 2019, whitehat hackers targeted two popular models of a car vendor that were manufactured before 2018 and exposed two vulnerabilities that could potentially cause denial-of-service (DOS) attacks. It is high time to study the cyber threats against connected vehicles and equip future connected cars with cost-effective automotive cybersecurity measures.
Figure 1: Cyber Vulnerabilities in Autonomous Vehicles
The above figure depicts the security issues perceived in autonomous vehicles. The most common attack pattern is Drive-By-Download which occurs when the vehicle gets exposed to the Internet. A malicious piece of code gets downloaded when the vehicle user accidentally sees or downloads any compromised website. The hacker will then be able to see sensitive data like passwords and user settings. Then the hacker can use any of the apps like Google Maps and can spoof real traffic or external sensor data, making the vehicle’s decision-making difficult. A new report released in August 2019 warned that there is a risk of autonomous cars being taken over by a hostile hacker. The US consumer group watchdog report said that if a cyber-terrorism attack were to take place on internet-connected cars during busy hours in major metropolitan areas of the US, then many people could be killed. This was demonstrated by scientists using CARLO, an autonomous vehicle simulator platform.
To establish a full-fledged autonomous driving experience, some core requirements are yet to be fulfilled. Fault tolerance is an absolute necessity considering the limited computing capacity of smart sensors and limited human intervention. Autonomous vehicles create a lot of data. The majority of data is unstructured, which necessitates the use of strong analytics to generate actionable data. Given that cloud computing is no longer a viable option due to its remote position, infotainment services can be efficiently offered in cars employing edge computing and caching. However, building and developing algorithms that can efficiently provide entertainment services in autonomous automobiles using edge computing and caching technologies is extremely difficult. The heterogeneous nature of firmware, devices, and data that the autonomous vehicle produces also poses a challenge to understand the contextual information of the overall application.
The nature of roads, electric vehicle charging stations, ethics, and the mindset of people willing to adapt to automated vehicles are the different hurdles that have to be crossed to realize the dream of the autonomous vehicle industry. On December 3, 2021, Union transport minister Nitin Gadkari had suggested the use of artificial intelligence for the enforcement of traffic rules and managing traffic congestion in India at the recent “all.ai” 2021 summit organized by Intel. Well, with all the green flags up and the global market for autonomous cars expected to reach US$45.05 billion by 2026, shouldn’t we make our vehicles smart enough to defend themselves from cyber-attacks?
- Union minister Nitin Gadkari wants artificial intelligence to be used for improving road safety | Autocar India
- Malik, S., and Sun, W. (2020). Analysis and Simulation of Cyber Attacks Against Connected and Autonomous Vehicles. 2020 International Conference on Connected and Autonomous Driving (MetroCAD). doi:10.1109/metrocad48866.2020.0
- Rasheed Hussain, and Sherali Zeadally (2018). Autonomous Cars: Research Results, Issues and Future Challenges, IEEE COMMUNICATIONS SURVEYS AND TUTORIALS, VOL. XX, NO. XX, 201X
- Autonomous Cars Market Size, Share, Trends, Opportunities & Forecast (verifiedmarketresearch.com)