In a significant advancement, researchers at TU Wien have developed a robot capable of learning how to clean washbasins by imitating human movements. This novel technology could revolutionize various industries, offering a glimpse into the future of autonomous robots.
Researchers at TU Wien have developed a pioneering robot capable of learning to clean a washbasin through human demonstration, a breakthrough that could have far-reaching implications across multiple industries. This novel approach to robotics involves a human guiding a specially equipped cleaning sponge along the perimeter of a sink, enabling the robot to observe and learn.
“Capturing the geometric shape of a washbasin with cameras is relatively simple,” Andreas Kugi, a professor at TU Wien’s Automation and Control Institute, said in a news release. “But that’s not the crucial step. It is much more difficult to teach the robot: Which type of movement is required for which part of the surface? How fast should the motion be? What’s the appropriate angle? What’s the right amount of force?”
Unlike traditional methods that rely on predefined mathematical formulas, TU Wien’s robot learns through mimicking human behavior. This strategy was showcased through multiple demonstrations where a human used a sensor-equipped sponge to clean the edge of a sink. By processing vast amounts of data from these demonstrations, the robot understands the intricacies of cleaning.
“We generate a huge amount of data from a few demonstrations, which is then processed so that the robot learns what proper cleaning means,” Christian Hartl-Nesic, head of the Industrial Robotics group in Kugi’s team, said in the news release.
The innovative learning algorithm, highlighted at the prestigious IROS 2024 conference in Abu Dhabi, has potential uses far beyond cleaning. The technology is applicable in various surface treatments like sanding, polishing and painting, making it a versatile solution for industrial tasks.
“The robot learns that you have to hold the sponge differently depending on the shape of the surface, that you have to apply a different amount of force on a tightly curved area than on a flat surface,” added Christoph Unger, a doctoral student from the Industrial Robotics group.
The vision for this technology extends beyond standalone robots. According to Kugi, these robots could be part of a larger network, sharing learned parameters and improving collectively through a concept known as “federated learning.”
“Let’s imagine many workshops use these self-learning robots to sand or paint surfaces. Then, you could let the robots gain experience individually with local data. Still, all the robots could share the parameters they learned with each other,” Kugi added.
The TU Wien team has already conducted extensive tests proving the robot’s flexibility and adaptability. With its debut at IROS 2024, where it received the Best Application Paper Award, this innovation is drawing international attention and acclaim, heralding a transformative era in robotics.