Researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed a new algorithm that allows a robot enough motion to give assistance to someone putting on clothes.
The team tested the robot by having it help put a jacket on a human, but the long-term potential of the robot could be to help those with disabilities or limited mobility.
“Developing algorithms to prevent physical harm without unnecessarily impacting the task efficiency is a critical challenge,” said Shen Li, MIT PhD student who led the research. “By allowing robots to make non-harmful impact with humans, our method can find efficient robot trajectories to dress the human with a safety guarantee.”
The MIT team started by determining how to properly model human movement, reactions and response to enable robot motion during human-robot interactions. A robot can attain fluent interaction if the human model is perfect, however, there is no flawless blueprint in humans.
If a robot was shipped to a home and needed a flawless blueprint, it would have a very narrow model to follow. Tasked with dressing, this become even harder as the robot would be based solely on the default model and bump into a human, possibly injuring them.
The MIT algorithm instead allows the robot to understand many possible models and mimic how a human can understand other humans. And as the robot gathers more data about its subject, the better it refines these models.
“This multifaceted approach combines set theory, human-aware safety constraints, human motion prediction, and feedback control for safe human-robot interaction,” said Zackory Erickson, assistant professor in The Robotics Institute at Carnegie Mellon University. “This research could potentially be applied to a wide variety of assistive robotics scenarios, towards the ultimate goal of enabling robots to provide safer physical assistance to people with disabilities.”