Have you ever picked up a closed box and moved it around or lightly shaken it to try and identify what was hidden inside? Well, thanks to AIM-SI hire Peter Yichen Chen and colleagues, this is no longer a skill exclusive to humans. Peter Yichen Chen, Chao Liu, Pingchuan Ma, Jack Eastman, Dylan Randle, and Yuri Ivanov have developed a method in which robots gently shake an object and through their internal sensors they are able to learn about an object’s weight, softness, or contents. While there are already methods that can do this through the use of computer vision, they are more complex, more expensive, and are not more accurate in determining mass. This technique is not only low cost, using only a robot’s internal sensors rather than additional cameras or measurement tools, but it is also more robust, allowing applications such as sorting objects in a dark basement or clearing rubble in a disaster zone.
The key to their approach is a simulation process that incorporates models of the robot and the object to rapidly identify characteristics of that object as the robot interacts with it. The researchers’ method leverages proprioception, which is a human or robot’s ability to sense its movement or position in space.
To learn more about this system and how it works, please see the original article.