“The robot can learn from corrective human feedback in order to plan its actions that are suitable to the environment and the objects present.”
Cornell University engineers have taught a robot to work in a mock-supermarket checkout line, modifying a Baxter robot from Rethink Robotics in Boston to “coactively learn” from humans and make adjustments while an action is in progress.
“We give the robot a lot of flexibility in learning,” said Ashutosh Saxena, assistant professor of computer science. “The robot can learn from corrective human feedback in order to plan its actions that are suitable to the environment and the objects present.”
Saxena’s research team will report their work at the Neural Information Processing Systems conference in Lake Tahoe, Calif., Dec. 5-8.
Modern industrial robots, like those on automobile assembly lines, have no brains, just memory. An operator programs the robot to move through the desired action; the robot can then repeat the exact same action every time an object goes by.
But off the assembly line, things get complicated: A personal robot working in a home has to handle tomatoes more gently than canned goods. If it needs to pick up and use a sharp kitchen knife, it should be smart enough to keep the blade away from humans.
The Baxter’s arms have two elbows and a rotating wrist, so it’s not always obvious to a human operator how best to move the arms to accomplish a particular task. So Saxena and graduate student Ashesh Jain drew on previous work, adding programming that lets the robot plan its own motions. It displays three possible trajectories on a touch screen where the operator can select the one that looks best.
Then humans can give corrective feedback. As the robot executes its movements, the operator can intervene, manually guiding the arms to fine-tune the trajectory. The robot has what the researchers call a “zero-G” mode, where the robot’s arms hold their position against gravity but allow the operator to move them. The first correction may not be the best one, but it may be slightly better. The learning algorithm the researchers provided allows the robot to learn incrementally, refining its trajectory a little more each time the human operator makes adjustments or selects a trajectory on the touch screen. Even with weak but incrementally correct feedback from the user, the robot arrives at an optimal movement.
Go deeper with Bing News on:
Flexibility in learning
- EDITORIAL: Fight for funding flexibility, not class-size reductionon April 11, 2021 at 9:00 pm
As a consequence, when students were learning in person, the classes were much smaller ... Education advocates should push for more school-specific flexibility with current funding, not more ...
- California Governor Newsom halts COVID-19 tiered restriction system, plans full reopening in June as deaths hit 60,000on April 11, 2021 at 8:41 pm
Against the backdrop of a growing fourth wave of the pandemic and new COVID-19 variants, California Governor Gavin Newsom declared he would fully reopen the economy in a little more than two months.
- Catholic educators share unexpected lessons of teaching in pandemicon April 11, 2021 at 2:52 am
Pandemic learning impacted Catholic preschools to high schools across the country and its success seemed to hinge primarily on the flexibility of students and teachers alike. That’s why it’s ...
- State gives schools flexibility to promote, graduate students without test scoreson April 10, 2021 at 6:19 am
Read more: About 44,000 Palm Beach County students must return to school for state exams More: Palm Beach County public schools to end distance learning this summer ... “This is the ultimate ...
- Portsmouth schools set date for full-time in-person learning to resumeon April 9, 2021 at 10:48 am
A joint message from Portsmouth school leaders states Gov. Sununu’s order allows the two extra days for the transition.
Go deeper with Google Headlines on:
Flexibility in learning
Go deeper with Bing News on:
- Gamma rhythm communication between entorhinal cortex and dentate gyrus neuronal assemblieson April 4, 2021 at 5:01 pm
Fernández-Ruiz et al. examined gamma-frequency activity and spike coupling between the entorhinal cortex and hippocampal dentate gyrus during learning and after selective perturbation of ...
- Husker center Cam Jurgens' next challenge: 'Ultimate control' of not only himself, but OL room, tooon April 3, 2021 at 1:00 pm
“Just be that voice during games and in practice and be consistent with it.” Just like learning to be a Big Ten center, developing into the voice of an entire position group is not an ...
- An emergent neural coactivity code for dynamic memoryon March 28, 2021 at 5:00 pm
Contingency-discriminating patterns were reactivated offline after learning, and their reinstatement predicted trial-by-trial memory performance. Moreover, optogenetic suppression of inputs from ...
- Balancing Democratization of Technology and Employee Collaborationon February 18, 2021 at 6:32 pm
The same goes for machine learning and cloud, as they both have compelling applications in people’s personal lives as well as at work. Besides changing the mindset of IT to designing open ecosystems ...