Furfaro Discusses AI in Space with Arizona Public Media

In a recent episode of Arizona Public Media's Arizona Science podcast, SIE professor Roberto Furfaro discussed his research into how artificial intelligence and “deep learning” can make automated space exploration vehicles more efficient. Furfaro explained how his work as the systems engineering lead for the Science Processing and Operations Center of the OSIRIS-REx Asteroid Sample Return Mission correlates with developing autonomous space vehicles. Artificial intelligence could make these vehicles more efficient, as well as allowing them to autonomously land on the moon, asteroids or Mars. 

"A few years back we published papers related to a challenging problem like, 'Can I design an intelligent system that can autonomously reason like planetary scientists?'," Furfaro said. "It's very hard to interpret data in real-time, especially when you don't have time to stream back the data for interpretation. If you go to places like Titan, for example, you're talking about 90 minutes just one way. This is extremely important for OSIRIS-REx to try to grab a sample from a very rough surface in a very tight spot."