Switching from one smartphone to another is generally a simple procedure. You sign in to your accounts and your apps, preferences, and contacts should sync with the new hardware. But in the world of robotics, swapping out an old robotic arm for a newer model has meant setting everything up from scratch.
To solve this problem, a team of researchers at Switzerland’s Federal Polytechnic School of Lausanne (EPFL) has developed what they call Kinematic Intelligence, a framework that makes switching robots function more like smartphones. They describe their system in a recent Science Robotics article.
Demonstrating skills
For years, robotics experts have been working to get robots to learn from demonstrations, teaching them new skills by showing them what to do, rather than writing lines of code. The idea is to remotely control or physically guide the robot’s arm to teach it a task such as cleaning a table, stacking boxes or soldering a car component. The problem is that most of these taught skills end up tied to the specific robot the training was done with.
But robotics is advancing rapidly. “Robots have different designs and new designs are being proposed today, which comes with its own set of challenges,” said Stithpragya Gupta, a roboticist at EPFL and lead author of the study. If a new robot has slightly longer links, a different joint orientation, or a more complex configuration, that learned behavior is instantly broken and the new robot will probably shake, freeze, or crash if it tries.
“New designs come with different capabilities and limitations,” said Durgesh Haribhau Salunkhe, a roboticist at EPFL and co-author of the study. “The problem is adapting to these limitations and capabilities, faithfully replicating the actions demonstrated by a human being.” Nowadays, making the jump from one robot body to another often means starting from scratch and retraining the entire system.
