Blue Sky’s the Limit
But current applications represent only a glimmer of this technology’s full potential, according to Yisong Yue. “Machine learning is going to transform our lives in ways that people cannot even imagine right now,” he says.
Yue should know: As an assistant professor in the Department of Computing and Mathematical Sciences at Caltech, he is part of the vanguard of researchers who are helping to usher in this transformative future.
Mind-Body Wholeness—for Robots
As a member of Caltech’s Center for Autonomous Systems and Technologies (CAST), Yue collaborates with engineers and scientists from many disciplines to advance research on autonomous systems such as drones and mechanical first responders. “CAST is about combining body and mind in robots,” Yue says. “I want to put the mind into the body.”
Caltech launched CAST in 2017 thanks to a generous endowment gift from Foster and Coco Stanback. Since then, corporations and other philanthropists have joined with Caltech to speed the development of next-generation robotics and machine-learning systems.
“Philanthropy enables the kind of speculative, blue-sky research that could pan out to be something really incredible,” Yue says.
Whole > Sum of Parts
Yue also is exploring ways to apply machine learning—and artificial intelligence (AI) more broadly—to other fields of study. At Caltech, this approach of taking computational thinking and combining it with another discipline to create something new has a long history and is known by the shorthand “CS+X”. For example, Yue is collaborating with Joel Burdick, Caltech’s Richard L. and Dorothy M. Hayman Professor of Mechanical Engineering and Bioengineering and a JPL research scientist, to develop a treatment that can incorporate machine learning to help patients with spinal injuries stand again.
Another project Yue is excited about involves exploring how he might apply his expertise to help aid in the design and manufacture of aircraft. “Right now, aircraft are designed using a combination of human ingenuity and mathematical modeling,” Yue says. “Perhaps we can use machine learning and AI to design an aircraft that has better aerodynamics and improved fuel efficiency.”
But this project presents Yue with a thorny question: If we use machine learning to boost the development of ingenious designs for aircraft, how can we prove that the resulting designs will be safe in the real world? Yue is working to combine machine learning with control theory, which models dynamics and stability. He hopes that engineers ultimately will use control theory to analyze machine-learning-designed aircraft and verify that they will be stable in flight.
When it comes to tackling these and other challenges, Yue believes that Caltech’s close-knit community—there are only 300 faculty members and 2,200 students—gives it a distinct advantage. “You’re more likely to interact with people outside of your immediate area of expertise,” Yue says. “What this means is that there are increased opportunities to find interesting connections between different fields. You get this potential for the whole to be much greater than the sum of its parts.”