AdaSky build thermal camera for self-driving car

AdaSky announced it has developed a far infrared (FIR) thermal camera and algorithm for detecting humans or objects on the road. Startup headquartered in Yokne'am Illit, Israel, designed camera system called 'Viper' to be applied to autonomous vehicles as a solution in which other sensors may have failed.

Tinuku AdaSky build thermal camera for self-driving car

AdaSky says the system will solve some of the complicated problems faced by self-driving cars by using thermal cameras as 'edge cases' to detect the heat of pedestrians, animals and other objects.

"Automakers are now beginning to understand that the current equipment is not enough, they need new sensors to help them with edge cases," said AdaSky's Vice President of R&D, Eyal Madar.

The automakers have been experimenting in recent years using cameras, radar, and Lidar to detect the movement of the surrounding situation. Each has the strengths and weaknesses in which all elements must communicate with each other to complement.

Thermal cameras are mature technologies originally developed for military use. Madar says AdaSky focuses on applications for autonomous vehicles, unlike larger FLIR Systems competitors.

"The last four to five years this technology has become available and cheaper," Madar said.

Thermal cameras are claimed to outperform traditional cameras during nighttime or fog and rainy weather conditions. Madar says the complete prototype will be ready in a year and mass production two to three years later.

Another challenge for autonomous vehicles is the inability to distinguish between images and the real thing. Sculptures or kangaroo images can be translated by the car as an actual kangaroo causing the car to brake.

The thermal camera will detect the temperature difference emitted and translate as a living object or inanimate object. Madar said AdaSky technology is capable of determining a living object at a distance of 200 meters.