June 18, 2014
Imagine a swarm of robots, whizzing through the air like a flock of birds, deployed to search or monitor an area, or, as Amazon CEO Jeff Bezos envisions, delivering your online packages directly to your front door.
It's a scenario that Behcet Acikmese is familiar with because he is working to make it, as well as other futuristic possibilities, reality. Acikmese, a UT aerospace engineering and engineering mechanics professor, researches the algorithms that inform the guidance and control systems in unmanned autonomous vehicles (UAVs)—a class of robots that includes space probes, driverless cars and military drones, among other devices.
"We will have a lot of things flying around in the future, a lot of things that are done by robots," Acikmese said, giving examples of UAVs fighting forest fires and exploring comets. "That will make our lives much easier, much more convenient...and safer, too."
He recently received three major grants for his UAV research: A $622,000 grant from the Defense Advance Research Projects Agency (DARPA), a $405,000 grant from the Office of Naval Research (ONR), and a $79,000 grant from NASA's Jet Propulsion Laboratory's (JPL) Strategic University Research Partnerships program (SURP). Acikmese is among the first researchers from UT to receive a grant from the SURP program, which is meant to foster collaborative research and educational activities between JPL and university partners. Paul Shapiro, a professor in the astronomy department, also received a SURP grant.
The grants fund a diverse variety of UAV research, including basic and applied. The DARPA grant is specific, with funds being applied to developing a decentralized control system for autonomous vehicle swarms, or a sort of 'hive mind' that directs the response of individual robots based to achieve a desired collective swarm behavior. The ONR grant, on the other band, funds basic exploratory research in mathematics related to path planning and optimization.
"It's not geared toward any application per say, but that's great because we get this funding for three years where we will explore techniques that will enable the solution of very complex decision making problems in real time," Acikmese said. "That's why I like it. This is real hard core, fundamental work that belongs in a university."
The SURP grant is on a middle ground between basic and applied. It funds research on algorithms that can help guide spacecraft adjust trajectory while conserving resources. Spacecraft guidance is a realm that Acikmese has particular expertise in, having developed the algorithm that directed the Curiosity rover's Mars landing in 2012.
At the root of all of Acikmese's current research is the basic problem of autonomous control and guidance—how to get a pilotless vehicle to automatically readjust its trajectory or objectives depending on ever changing conditions. To solve this problem Acikmese invented a mathematical method called "lossless convexification" that improves UAVs' ability to make more precise trajectory decisions by retaining data that in past methods was lost during the computing process, hence the "lossless" descriptor.
"If you can [successfully compile various factors], you reach what's known as 'convexification.' Problems can then be solved with robust, reliable, very, very fast numerical methods," Acikmese said. "They are the most reliable optimization methods that we have."
The method plays a key role across the grant-funded research. It's also a method that Acikmese says could enable future space probes to better target landing sites, allowing future planetary probes to touch down in canyons and craters.
Currently, high-tech UAVs are found most often in the government realm. But Acikmese says it's only a matter of time until they branch out into the wider consumer market—opening up new opportunities for aerospace students in their wake.
"Amazon.com, one of the biggest companies in the world, is investing in UAVs. Google is buying a lot of robotic companies. It's a new revolution coming," Acikmese said. "I think it's exciting. I wish I was a student at this time."
