Sign In

Communications of the ACM

ACM TechNews

MIT Robot Combines Vision, Touch to Learn Game of Jenga


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
The robot playing Jenga.

Engineers at the Massachusetts Institute of Technology have created a robot that taught itself to play Jenga by working out the best way to execute tasks based on both visual cues and physical interactions.

Credit: Alberto Rodriguez et al

Massachusetts Institute of Technology (MIT) engineers have created a robot that learned to play Jenga by working out the best way for executing tasks based on both visual cues and physical interactions.

The robot has a soft-pronged gripper, a force-sensing wrist cuff, and a camera, used in combination to visualize and feel the Jenga tower and its individual blocks.

As the machine pushes against a block, a computer digests visual and tactile feedback from its camera and cuff, and compares these measurements to previous moves. It also considers outcomes—whether a block, in a certain configuration and pushed with a certain amount of force, was successfully extracted or not.

Said MIT's Alberto Rodriguez, "The key challenge is to learn from a relatively small number of experiments by exploiting common sense about objects and physics."

From MIT News
View Full Article

 

Abstracts Copyright © 2019 SmithBucklin, Washington, DC, USA


 

No entries found