|
 |
Costello over Bagram Air Base Afghanistan 2013 piloting an EA-18G. |
|
As drones, self-driving cars, and other autonomous vehicles revolutionize transportation, the need to establish frameworks for reliable, safe operation is paramount. Donald “Bucket” Costello, the newest member of the University of Maryland’s (UMD) aerospace engineering faculty, focuses much of his research on addressing this need.
Costello joins the department after already having headed up testing and evaluation of autonomous systems at the A. James Clark School of Engineering’s MATRIX Lab in Southern Maryland, where he established a graduate program in the field. The first two graduate students from the program, Jacob Safeer (MS AE) and Ryan Lowe (M.Eng, Robotics) received their degrees at commencement this spring—along with Costello, who completed his doctorate in mechanical engineering at UMD in 2020 but missed the chance for an in-person commencement due to the COVID pandemic.
Dr Costello with his wife Jenna, son Jack, and daughter Finley.
Before joining UMD he retired from the United States Navy, having served 25 years as a naval aviator, test pilot, acquisitions professional and military professor. His last tour was at the United States Naval Academy where he was an assistant professor in the Weapons, Robotics, and Control Engineering Department and served as associate chair.
“My current research focuses on finding ways to enable systems without a human in or on the loop to operate in our everyday lives,” Costello said. “This is especially relevant to autonomous drones. How does a drone operate without human involvement and oversite? How do we make a risk-based decision to allow that to happen, without ultimately having a human to blame when things go wrong?”
To answer questions like these, Costello examines methods of drone perception—that is, the ways in which drones can develop situational awareness of their environments. Perception goes hand-in-hand with decision-making: a drone must not only be able to sense its surroundings, but act based on those inputs. Engineers also have to set parameters and limits governing the range of actions it can take.
For example, a drone might notice that it is on a collision course with another aircraft, forcing it to maneuver out of harm’s way. Engineers would need to ensure that the maneuvers don’t result in other problems, such as violating Federal Aviation Administration regulations or inadvertently crashing the drone.
Costello employs a variety of methods in his research, from computation to modelling and simulations, as well as real-world tests. With support from the Office of Naval Research, he will soon take over as “air boss” of a 108-foot naval vessel operated out of the Naval Academy, with the goal of integrating autonomous platforms that do not require control by the ship’s crew.
His current work extends research he conducted as part of his doctoral thesis, completed in 2020 under the direction of UMD Aerospace Engineering Professor Mumu Xu. Costello was on active duty at the time, serving on board ships and calling in from various ports to update Xu on his progress.
It’s also the logical extension of a lifelong passion for flying that began while he grew up around naval aviation in San Diego, a period that coincided with Top Gun’s 1986 theatrical release.
In fact, along with his academic expertise, Costello brings to the UMD Aero faculty a credential worthy of the fictional Maverick—his name is inscribed on an EA-6B Prowler jet now on display at the National Naval Aviation Museum in Pensacola, Florida. He dropped it off during its final flight in 2010 while he was serving as a Developmental Test Pilot at Naval Air Station Patuxent River.
“I’ve been fortunate in being able to achieve many of my childhood dreams, including flying jets for the Navy, raising two wonderful children, and spending my life with my incredible wife,” Costello said. “Now my family and I are excited for the opportunity to fulfil another dream of mine: to become a tenure-track faculty member at my alma mater!”
May 22, 2026
|
