April 1, 2019 at 19:25 #2524
A lot has been going on with autonomous vehicles these days. Almost every famous car manufacturer seems to be in a rush to master the perfected version of self-driving cars. While most people believe that fully autonomous vehicles are still in their infancy stage, manufacturers think that they will dominate the roads by 2020.
The problem with autonomous cars is not the manufacturing process itself but the volume of probable outcomes that could arise from even the tiniest of miscalculations. However, governments and manufacturers have more to worry about than just manufacturing defects. What really bothers us – when it comes to traveling in self-driving vehicles is their security, i.e., the ability to ward off cybercriminals.
As the Internet of Things (IoT) is an embedded feature that can help cars of the future navigate through busy roads, making split-second decisions, the risks of falling prey to a malicious hacker is likely the most significant concern manufacturers will face going forward.
An autonomous car is already facing growing controversy over its safety and ethical decision-making capabilities. Critics have warned that self-driving vehicles would fail ethical testing in various scenarios. Some autonomous cars during the testing phase got involved in collisions and made poor moral decisions. Hence, there is a growing concern over structuring policies and legislation to accommodate legal action against cars that can be driven independently.
With a vehicle that is programmed to increase passenger safety and improve road safety in general, what happens when such cars make mistakes? Who should be held accountable for a collision in such cases? How would an autonomous vehicle react when it has to choose between the passenger’s safety and the pedestrians? All these questions have raised eyebrows for many. However, that is not the only challenge manufacturers’ face when it comes to self-driving vehicles.
One possible challenge that is widely discussed is whether these vehicles will have fortified cybersecurity safeguards to keep cybercriminals at bay. Every major company is competing to manufacture the most exceptional autonomous car. It is almost like a race, but these cars will fail to deliver the results if security concerns aren’t tackled. It is vital for manufacturers to identify the vulnerabilities that bad actors can manipulate to access the vehicle’s mainframe.
Here are some ways self-driving cars are susceptible to hackers.
As we all know, self-driving vehicles process immense amounts of data that is stored and retrieved using cloud computing. The car sends its GPS location and obtains information from other automobiles to predict traffic flow and to distinguish the best route between places. Also, AVs need to transmit and retrieve data in real-time from surrounding cars for a smooth and safe driving experience.
With so much data processing involved, some vulnerabilities can be exploited by malicious hackers. For example, a malicious hacker would only need to hack one entry-point, i.e. the Cloud-Computing database or the vehicle’s communication device, to access large amounts of data that can be used to manipulate the car.
A bad actor could also willingly switch off safety features to turn a perfectly safe vehicle into a problematic one. Similarly, with the need for constant information transfer, the information cannot be highly encrypted, or it would delay the transmission of data, which in itself can be damaging for the passenger.
MULTIPLE CODING LANGUAGES
A single company does not manufacture the components used in vehicles these days. In fact, many elements are purchased from other manufacturers, as it saves time and money. Each manufacturer would have a different coding system that needs to be aligned to make information transmission compatible with other components.
As there are too many variables involved in different components interacting with each other, identifying a weak link could be a little more than a challenge, which manufacturers might overlook. For cybercriminals, this would be the perfect opportunity to exploit security weaknesses.
A quick way around this problem is penetration testing. Pen testing is a systematic process to identify vulnerabilities and weaknesses in applications. It is a controlled form of hacking to unveil flaws in a system that is used by the manufacturer for reinstating security features. During a penetration test, the tester could mimic the actions of hackers and look for susceptibilities in the vehicle but without causing havoc. This would allow manufacturers to strengthen their anti-hacking software and make autonomous driving vehicles more resilient to hackers.
COMBINING THE TECHNOLOGY AND RESOURCES
Car manufacturers are in somewhat of a hurry to present the world with their first perfect autonomous vehicle. In all fairness, the heightened competition could be the potential cause of delay in the launch of self-driving vehicles. What car manufacturers have failed to demonstrate is a collaboration where they share their technology and resources to build a car that is more refined and secure.
Some of the most renowned companies, such as Toyota, Tesla, and Google, are investing millions of dollars in creating the ultimate self-driving car. The result is slow progress in the domain of autonomous vehicles.
Since manufacturers are not willing to share their technologies, it would only give hackers more opportunities to exploit self-driving vehicles. Shared resources and information would be an excellent solution for tackling cyber-security implications in autonomous cars.
Collaboration between two manufacturing companies would mean better resources to develop a resilient technology that can secure the cars’ infrastructure and keep hackers locked out. Shared resources would also improve testing the vehicle for vulnerabilities, as there would be a more exceptional workforce at the disposal of the companies. Therefore, manufacturers need to collaborate with each other instead of competing to improve cybersecurity implications in autonomous vehicles.August 31, 2019 at 18:45 #5390
You must be logged in to reply to this topic.