July 7, 2016
A team of University of Zurich scientists recently taught a robot how to track and hunt another, human-controlled robot. The results are expected to offer valuable data that could help future robotics assess environments in order to track subjects in real time.
Robot tracking isn’t quite as horrifying as it sounds, according to one scientist on the team, which is from the U of Z’s Institute of Neuroinformatics. As professor Tobi Delbruck explains it to Motherboard, “Following [self-driving cars or drones] is the obvious application, but one could imagine future luggage or shopping carts that follow you. This way, the problem is less like a predator and its prey and more like herding, or a parent and child.”
How Robot Tracking Works
In addition to an ever-useful deep learning neural network, the robot works with a “silicon retina,” which is designed to track moving targets better than a mere camera:
“If a robot tracked its prey with a regular camera, the slow frame rate would result in a series of frames that didn’t really represent the true path of movement, especially if the escaping prey were moving quickly. The silicon retina, instead of producing frames, contains pixels that individually and autonomously detect changes in illumination and transmit that information in real time, which results in a steady flow of visual information instead of a series of disjointed images.”
Advancements in robotics are already helping plenty of fields, from smartphones to healthcare, but a way for robots to track subjects in real time is a function with applications across even more sectors. Robots moving around autonomously in the world need to be able to figure out how to follow things.
Did you like this article?
Get more delivered to your inbox just like it!