Helping robots zero in on the objects that matter

MIT engineers have developed Clio, a method enabling robots to make task-relevant decisions about their surroundings based on natural language descriptions. In experiments, Clio successfully identified and mapped only relevant parts of various environments, from apartments to office buildings. The system runs in real-time, allowing robots to quickly survey areas and complete tasks while ignoring irrelevant objects. Researchers believe Clio could be useful for search and rescue, domestic robots, and factory settings.