September 30, 2011
Whether in the sea or in the air or on the ground - on this planet or another - if the vehicle runs, Dr. Maruthi Akella is interested in studying it and improving it.
"I keep telling students that anything that runs and moves is of interest to me," Akella says with a grin. "I'm agnostic as to the terrain."
The full range of applications for Akella's research in guidance and control runs the gamut: from deep space to lower orbit just a hundred kilometers from earth surface to atmospheric flight to terrestrial robots and underwater vehicles, like the ones Akella collaborated on with the U.S. Navy's Space and Naval Warfare Systems Center.
Recently, Akella has directed his research focus on two big projects: one which provides NASA with a cheaper, faster, efficient alternative to the high-grade tactical navigation sensors they currently use on their vehicles, and another which looks into autonomy, responsiveness, and hazard avoidance for vehicles in orbit or in space.
The first involves Project Morpheus, a NASA-designed vehicle engineered at Johnson Space Center which is used to test integrated propulsion, guidance, navigation and control systems on flights to the moon and Mars. Whereas NASA has traditionally used high-grade tactical stand-alone navigation sensors on vehicles, Akella's research team is investigating using swarms of smaller, cheaper, and commercially available sensors which communicate wirelessly with one another and function as efficiently and accurately as a group as the expensive stand-alone sensors do. The current sensors NASA uses cost half a million to three quarters of a million dollars and take over a year to be ordered, delivered and tested for meeting the stringent flight standards, a process Akella said simply isn't amenable to the rapidly moving world we live in today.
Compare that to a couple dozen off-the-shelf navigation sensors that cost about $80 each and function just as well.
"The overall performance when they're all integrated as a swarm of 20 or 30 units is compatible to these tactical sensors we've been using," Akella said. "The idea is that sensors and resources within the swarm can collaborate and share information. They work as a team and ultimately give you a good coherent navigation solution to enable precision space maneuvers."
The cheaper sensors are just a couple inches in diameter, can be easily replaced, and provide plenty of redundancy. A prototype sensor swarm built at UT will fly with Morpheus in spring 2012 as a backup system, but the ultimate goal is that these swarms will replace the expensive, larger sensors. The test flight is also an opportunity to validate their theoretical calculations: "We know the performance of each unit and how they're talking to one another to find out how many we need so their combined performance is the same as what you'd get with a tactical flight grade sensor," Akella said.
The second project aims to solve an increasingly critical problem that has attracted the Air Force's attention as well: how best to maneuver for avoiding objects in space that interfere with a predetermined flight plan.
"Historically, all our resources in space, for the most part, are more or less sitting ducks," Akella said. "They're not really protected. If you have a very expensive satellite out there that's of national importance, you need situational awareness and responsiveness to defend the resource. It doesn't matter if it's an antagonistic object with the intent to destroy or it's just orbital debris tumbling toward your resource."
Adding autonomous hazard avoidance algorithms to satellites and space vehicles would offer a significant level of protection that doesn't exist right now.
"You have to be able to responsively discriminate the intent of space objects and then come up with a good way to maneuver without moving too far from where you were headed," Akella said. "You don't want to be running way from a truck and into a train." Autonomy also allows for changes in flight plans that might arise not only because of an unanticipated hazard but also because of unexpected opportunities.
"Currently, most flight planning and execution is rigid," Akella said. "You're constrained by what you already decided to do and things don't always work out the way you plan."
He provides a hypothetical example: "Say you want to go to the moon or to an asteroid or wherever NASA wants to go next. If you see some interesting feature at a landing site, we need really good onboard logic to help autonomously land there without much ground-support and without running into a big rock."
Whether it's pursuing transformative advances toward commercially viable swarm and cooperative robotic alternatives to the old way of vehicle navigation or developing autonomous guidance and control systems that allow for opportunistic sensing and more freedom for vehicles to explore, Akella's research keeps him at the cutting edge of the aerospace industry.
