Engineers at Massachusetts Institute of Technology (MIT) have developed an autonomous vehicle (AV) sensor system that identifies even slight changes in shadows, which inturn helps in detecting objects moving around corners.

According to MIT, AVs could one day use this system to avoid potential collisions with vehicles or pedestrians coming from around corners.

The researchers conducted successful experiments with an autonomous car in a parking garage and an autonomous wheelchair in a hospital.

According to the engineers, the car-based system identified the object more than half a second faster than traditional LiDAR, which only detects visible objects.

This short half-second could make a huge difference in cases of fast-moving vehicles.

The Andrew and Erna Viterbi professor of electrical engineering and computer science and Computer Science and Artificial Intelligence Laboratory (CSAIL) director Daniela Rus said:  “For applications where robots are moving around environments with other moving objects or people, our method can give the robot an early warning that somebody is coming around the corner, so the vehicle can slow down, adapt its path, and prepare in advance to avoid a collision.

“The big dream is to provide ‘X-ray vision’ of sorts to vehicles moving fast on the streets.”

At present, the new system has only been tested in indoor settings, where the speed of robotic systems tends to be much lower and where lighting conditions are consistent. These conditions make it easier for the system to sense and analyse shadows.

The researchers built the ‘ShadowCam’ technology using computer-vision techniques to find and categorise changes to shadows on the ground.

ShadowCam uses sequences of video frames from a camera targeting a particular area, such as the floor in front of a corner.

It then detects changes in light intensity over time, especially if something is moving. Some of these changes may be too invisible to the naked eye but gets analysed by the technology through various properties of the object and environment.

ShadowCam then computes this information and categorises each image as containing a stationary object or a dynamic one. When dynamic images are detected, it reacts accordingly.