Assuming you are confronted with the task of ascending a flight of stairs while carrying a sizable, weighty box. Your approach might involve splaying your fingers and employing a two-handed grip to hoist the box, subsequently positioning it atop your forearms. As you steady it against your chest, your entire body comes into play, orchestrating a coordinated effort to maneuver the bulk.
To address this, researchers at the Massachusetts Institute of Technology (MIT) have now developed an Al technique that enables a robot to create intricate plans for handling an object using the entire hand, instead of just the fingertips. On a typical laptop, this model can produce useful blueprints in approximately a minute.
The technique developed, named “Contact-rich Manipulation Planning” uses an AI technique called smoothing, which summarizes many contact events into a smaller number of decisions, to enable even a simple algorithm to quickly identify an effective manipulation plan for the robot.
H.J. Terry Suh, an electrical engineering and computer science (EECS) graduate student led the research paper alongside Tao Pang PhD ’23, a roboticist at Boston Dynamics AI Institute; Lujie Yang, an EECS graduate student; and senior author Russ Tedrake, the Toyota Professor of EECS, Aeronautics and Astronautics, and Mechanical Engineering, and a member of the Computer Science and Artificial Intelligence Laboratory (CSAIL). The research appears this week in IEEE Transactions on Robotics.
Contact-Rich Manipulation Planning: A New AI Technique for Robots for Manipulating Objects
This approach, albeit still in its infancy, may eventually allow industries to use smaller, mobile robots that can control objects with their complete arms or bodies rather than big robotic arms that can only grasp with their fingers.
Additionally, since they could swiftly adjust to their surroundings using only an onboard computer, this method could be helpful for robots sent on exploration missions to Mars or other solar system worlds. Hence, this might save prices and lower energy use.
“Rather than thinking about this as a black-box system, if we can leverage the structure of these kinds of robotic systems using models, there is an opportunity to accelerate the whole procedure of trying to make these decisions and come up with contact-rich plans,” says Suh, the led author.
Since the model they created is based on a cruder approximation of reality, it is unable to manage extremely dynamic motions, such items falling. Although efficient for more laborious manipulation tasks, their methodology is unable to produce a strategy that would allow a robot to toss a can into a garbage can, for example. The researchers want to improve their method in the future to be able to handle these extremely dynamic motions.
“If you study your models carefully and really understand the problem you are trying to solve, there are definitely some gains you can achieve. There are benefits to doing things that are beyond the black box,” Suh added.
