Listen to this article
Researchers at University of Washington have developed a tool capable of designing 3D-printable passive grippers so that robots can more easily switch between different tasks.
Many robots are tied to a single job and are unable to change gears if necessary to perform a task outside of their usual task. Researchers at the University of Washington hope to solve this problem by creating a system capable of designing passive grippers, so that the robot can switch grippers and complete a task with new objects.
To design the grippers, the team provides the computer with a 3D model of the object it is going to pick up and its orientation in space. The team’s algorithms then generate possible grip configurations and rank them based on stability and other metrics.
Then the computer selects the best option and co-optimizes it to see if an insertion trajectory is possible. If it doesn’t find one, it moves on to the next option until it finds one that will. When it does, the computer sends instructions to a 3D printer to create a gripper and to the robot arm to find the object’s grip path.
The researchers tested his system on 22 different objects and successfully selected 20 of them. For each form, the researchers carried out 10 detection tests. For 16 of the shapes, all 10 tests passed.
The robot was unable to select two of the objects due to problems in the 3D model of the objects that were given to the computer. The first object, a bowl, was modeled with thinner sides than they actually were, and the second object, a cup with an egg-shaped handle, was misoriented.
The system excelled at picking objects that varied in width or had protruding edges, and struggled with uniformly smooth surfaces, like a water bottle.
The research was funded by the National Science Foundation and a grant from the Murdock Charitable Trust. Authors of the article include Milin Kodnongbua, Ian Good, Yu Lou, Jeffrey Lipton and Adriana Schulz.