A group of Canadian researchers is seeking to refine how humans and robots collaborate with each other in industrial settings.
Industrial robots are usually “fixed and programmed to operate at high speeds and perform tasks such as welding, painting, assembly, pick-and-place and material handling,” states a Dec. 14 press release from the University of British Columbia, Okanagan. Social robots, meanwhile, are typically programmed to work in many different environments.
The researchers, using autonomous vehicle guidelines, have developed a system that combines artificial intelligence and machine learning to help guide robots. In part, that means the machines can predict what other robots and humans will do, then respond in kind.
“Increasing automation levels is standardized and accepted by the automotive industry, but other industrial settings, while relatively static, don’t have the same standards,” Debasmita Mukherjee, a doctoral student at UBC Okanagan and lead author of a soon-to-be published paper on robot learning strategies for human-robot collaboration in industrial settings, said in the release. “In the future, not only will industrial automated systems continue to use sensors to enable perception and communication similar to human capabilities, but they will also be adapting and communicating in real-time with their surroundings.”
The paper introduces “a novel taxonomy of levels of interaction between humans and robots” and features “a comprehensive review of the machine learning methodologies and industrial applications of the same in the context of adaptable collaborative robots.”
The researchers are hoping to develop systems that allow robots to “function and respond outside of a prescribed environment,” such as a factory. “The endgame is to achieve the seamless team dynamics and communication fluency of an all-human team while using robots.”
The paper is set for publication in the February edition of the journal Robotics and Computer-Integrated Manufacturing.