MIT develops a new way to make chip sets for robots

MIT develops a new way to make chip sets for robots   | Photo Credit: MIT

(Subscribe to our Today's Cache newsletter for a quick snapshot of top 5 tech stories. Click here to subscribe for free.)

MIT researchers have developed a new way of customising chip sets for robots. Called "robomorphic computing", the method can speed up a robot’s response time and enable it to think faster.

The method helps build chips by using a robot’s physical layout and its intended application. The hardware design is based on user inputs, like a robot’s limb layout and how its various joints can move. It then translates these physical properties into mathematical matrices corresponding to a robot's movements.

Robots often don’t move quickly in complex situations, and “the hang up is what’s going on in the robot’s head,” said Sabrina Neuman, lead developer of the new process, and researcher at MIT’s Computer Science and Artificial Intelligence Laboratory.

Robots perceiving stimuli and calculating a response needs a “boatload of computation,” which limits reaction time. But the new way is said to fix the mismatch between a robot’s “mind” and body.

Neuman plans to automate the entire system of robomorphic computing so users just need to drag and drop their robot’s parameters, and “out the other end comes the hardware description."

“Ideally, we can eventually fabricate a custom motion-planning chip for every robot, allowing them to quickly compute safe and efficient motions,” said Brian Plancher of Harvard University, co-author of the paper ‘Robomorphic Computing: A Design Methodology for Domain-Specific Accelerators Parameterized by Robot Morphology’.

This article is closed for comments.
Please Email the Editor

Printable version | Mar 1, 2021 8:58:16 AM |

Next Story