Semi-autonomous driving system reduces accidents by 75% [w/video]

Scientists from MIT have developed a semi-autonomous driving system that is proven to reduce acccidents by 75 per cent.

A front-mounted camera and laser range-finder identify the car's surroundings while making realtime updates to distances, mapping out the car's movement into 'safe zones'. The system also asseses the vehicle's limits, such as grip, and corrects the car.

Over 1,200 trials have taken place using a Kawasaki Mule containing Light Detection and Ranging (LIDAR) technology and various other devices, showed a reduction of accident by 75 per cent. It's thought that minimizing the reader's blind spot would make it almost 100 per cent successful.

The system's developers envisage the system used by a dash-mounted smartphone that uses the camera as well as accelerometers and gyroscope to help prevent accidents on the roads, paving for the way for a future app.

Ph.D. student Sterling Anderson and Karl Iagnemma, principle research scientist at the MIT Robot Mobility Group, designed the algorithm with Quantum Signal LLC, a small technology developer in Michigan.

Of course autonomous driving isn't a new thing. Google has been leading the charge towards a fully-autonomous car, releasing a video in April showcasing its self-driving Prius. Its system takes traffic information and senses activity nearby to adjusts the car's behavior.

The latest Cadillac XT S includes sensors, cameras and radar that can provide the car with all the information needed to avoid crashes and to aid in parking and GM aims provide semi-autonomous cars by 2015 and fully-autonomous cars by 2020. Seemingly every other major manufacturer also has plans to follow.

By Rufus Thompson