How Easily Can Autonomous Vehicles Be Hacked?

• With robot vehicles slowly becoming more common, researchers are saying that hacking them is possible.
• Researchers successfully demonstrated three attack techniques that slipped past undetected by common detection techniques.

Introduction

Drones and rovers are popular robot vehicles that are starting to find many uses in several sectors including agriculture, package delivery, warehouse management, and more.

• Researchers designed stealthy attacks against these vehicles and discovered that these vehicles could be hacked quite easily.
• These attacks did not require any human intervention, and they worked on real and simulated robot vehicles.
• It is said that none of these techniques could be detected by the commonly used detection techniques.

More about the attack scenarios

The research team worked primarily on three attack scenarios.

• False data injection, that involves injecting minor deviations to the data over a period of time. The idea behind working on this attack technique is that malicious actors may be able to inject small deviations continuously over a period of time to divert the vehicle to a location of their choice.
• Artificial day, that involves changing the timings of essential actions in the process.
• Switch mode, which is a form of false data injection. This attack technique involves introducing malicious code into the system that will cause the vehicle to performs a mode different from the intended one.

Researchers have detailed the attack scenarios and have also outlined countermeasures for dealing with such attacks in the latest paper.

What they’re saying

“We saw major weaknesses in robotic vehicle software that could allow attackers to easily disrupt the behavior of many different kinds of these machines,” said Karthik Pattabiraman, a professor who supervised this research.

“Robotic vehicles are already playing an important role in surveillance, warehouse management and other contexts, and their use will only become more widespread in the future,” said the study’s lead author Pritam Dash.