Can an Ultra-Fast Camera Make Self-Driving Cars Safer?

Chen with a prototype of the Celex untra-fast camera. (Image courtesy of NTU.)

When futurists imagine the roadways of tomorrow, collisionless thoroughfares filled with fast moving streams of driverless cars are a common theme.

Though driverless capabilities are rapidly advancing, there’s still a good deal of work to be done before humans release the wheel. One area of driverless technology that needs improvement is the visual sensing and processing that guides ever vehicle, and researchers in Singapore believe they’ve made a big leap forward.

Named Celex, this improvement to driverless technology comes in the form of an ultra-fast camera that can see far ahead, resolve objects in low light and construct data-light images that can be processed rapidly.

In a normal driverless camera system, the imager collects 120 frames of video data per second. In most instances, these images can be obscured by varying light conditions and can be very data-intensive. Essentially, these two characteristics can prove dangerous for driverless passengers because the systems controlling a car are receiving faulty data that takes a long time to process. Between those two attributes, accidents can occur.

“Our new camera can be a great safety tool for autonomous vehicles, since it can see very far ahead like optical cameras but without the time lag needed to analyse and process the video feed,” explained assistant professor Chen of NTU’s School of Electrical and Electronic Engineering.

Chen’s Celex system gets around the pitfalls of traditional driverless imaging systems by recording the changes in light intensity between scenes at nanosecond intervals and using an onboard circuit to do instant analysis of these scenes. By leveraging quick, lightweight data the Celex system can deliver information in near real-time making it possible for an autonomous car to be more “aware” of its surroundings and more capable of safe driving.

“With its continuous tracking feature and instant analysis of a scene, it complements existing optical and laser cameras and can help self-driving vehicles and drones avoid unexpected collisions that usually happens within seconds.” Chen continued.

The Celex system is now in its final prototype stage and could be available near the end of the year. At the moment Chen has found a number of private companies that are interested in his technology, meaning that not only does he stand to make a bit a scratch, but the futurist vision of tomorrow’s roads might be picking up speed.

For a dissenting view, find out why Self Driving Cars Mean Less Than You Think.