
Satellite repair is an expensive undertaking for any business involved in operating them. New advancements in the field of AI and termed ‘genetic fuzzy logic’ offers hope for companies with satellites in orbit that may require repair or refueling. Ou Ma, professor at the University of Cincinnati, along with his research assistant Anoop Sathyan are working on a method to create robotic systems that can work together or independently to repair satellites while they’re still in orbit around the planet.
Repair Satellite Requirements
According to Professor Ma, the best repair satellites would be able to offer several different functions. The professor was previously involved with various projects dealing with robotic arms on the International Space Station. The experience provided by these projects suggested to him a need for autonomous robotic repair satellites that can work collaboratively or independently as the need arises.
Promise for the Technology
Teaching robots to work collaboratively is no easy task. The researchers devised a smart way of teaching robots how to respond to stimuli from other members of their group. By using a game where a token required to be put at a particular location, but could only be done so by controlling individual strings (each assigned to a specific robot), the team taught the robots about working in a group. By leveraging the AI technology, the robots could detect stimulus from other robots and respond to suit.
Saving Time and Money
Repairs on satellites are expensive because they require a human being to be sent up to perform fixes manually. Many devices develop defects after they’ve been put into orbit. A good example is the Hubble Space Telescope, which had a lens defect only discovered after the device was already circling the planet. NASA is already looking at launching a satellite in 2022 to intercept and refuel with another orbiting device as a proof of concept for autonomous repairs. The genetic fuzzy logic technology may offer the space agency options for their technology.