Where the world is waiting for self driving cars to become more popular to reach the masses, a security researcher has found a major flaw in the driverless car that can possibly drive it off the road.
Principal scientist at software security company, Security Innovation, Jonathan Petit, discovered that a laser pointer that costs only $ 60 could interfere with the laser ranging (Lidar) system of the car that could bring it to a halt.
Most self-driving cars rely on to navigate on this system of Lidar which creates a three dimensional map and allows the car to see potential hazards by bouncing a laser beam off obstacles.
Focusing the laser pointer at an automated or a semi automated car will be picked up by the Lidar system and can trick the car into thinking of some objects ahead it while there’s nothing actually. This act will force the car to slow down. A hacker can also overwhelm it with spurious signals which will force the car to remain stationary.
During his tests, Petit recorded laser pulses reflected by a commercial Lidar system, and then mimicked them with the laser back at the navigation system. This method worked from a distance of 300 feet from the car, and didn’t require perfect accuracy with the laser beam.
According to him, the movement of cars, pedestrians or stationary obstacles can be imitated from 50 to 1000 feet away from the car and the same attack can be carried out using a Raspberry Pi or an Arduino single-board computer.
On detecting a phantom object, the car may exhibit both short and long term response. The short term reaction may only consist of an unnecessary stop but a long term stop may trick the car into believing a blockage on the road thus taking an alternative route which will affect the trip.
The automakers need to ensure that simple hacks don’t render driverless vehicles useless or worse.
If proper steps are not taken on security implications of internet-connected cars right now, they will be vulnerable to hackers in the same way as PCs, laptops and tablets.
Director of smart connected vehicles at Cisco, Andreas Mai believes that an advanced end-to-end security reference architecture and close collaboration among automakers, suppliers, technology providers and government agencies should be maintained in order to deal with modern cyber attacks.
In a world, where data breaches takes place every time and all sorts of corporations look up to cyber security to protect their customer’s personal and financial information, car companies have something major to worry for.
Automated cars were developed with thought for safety as the conventional, human-driven cars produced many instances of bad decisions of humans while driving. Road accidents happen because of human errors on when to accelerate and when to put brakes.
But Google, which has led the way on self-driving cars, has experienced several accidents since hitting the road. In July, one of the firm’s Lexus SUV driverless cars was rear-ended in Google’s home city of Mountain View, California.
For car companies, the worry of hacking does not end with financial crimes and frauds like in other corporations but here hacking can result in real-world and real-time physical problems and injuries.
While automated cars could be beneficial in future, the companies that bring them to the masses have to make people comfortable about them. They won’t be successful if they aren’t perceived as completely safe.