|
Space is full of objects that are extremely hard to detect, including human-made debris as well as tiny asteroids. A new, algorithmic approach being developed by Tam Nguyen, assistant professor at the University of Maryland (UMD) Department of Aerospace Engineering, could help find and track such objects more quickly and efficiently than is currently possible, without the need to build larger telescopes or wait for better sensors.
Nguyen, a Massachusetts Institute of Technology graduate who joined the UMD faculty in 2024, recently received a Defense Advanced Research Projects Agency (DARPA) Young Faculty Award that will support her work. The award, which provides up to $250,000 over a base period of two years and an optional third year, is given to rising stars in junior research positions.
According to DARPA, the YFA program supports “innovative new research that enables transformative Department of Defense capabilities,” and is intended to help develop the next generation of researchers focused on national security issues.
Nguyen has built a search algorithm, dubbed the Fast X-Ray Transform, that uses a “divide-and-conquer” strategy, as she puts it, to search large areas of space.
Ultra-faint objects are hard to find, especially if their movements are unknown because they can’t be tracked long enough to achieve a sufficient signal-to-noise ratio. Researchers, however, can hypothesize object trajectories and then examine sensor data to see if it lines up with any of their hypotheses, in an approach known as “synthetic tracking” or "track-before-detect."
The time involved in examining multiple possible trajectories can pose a daunting barrier, however, especially when the search space is large. Nguyen’s algorithm cuts down on that time by honing in on short segments and then combining these segments into a larger search space.
It’s effective because possible trajectories often share at least some segments. Thanks to the algorithm, those shared segments don’t have to be examined again and again.
“The beauty of it is that you only have to compute these segments once, and then you have the data you need for many different trajectories,” Nguyen explains. “This turns out to be a very elegant solution if we’re looking for linear trajectories.”
Being able to detect elusive objects is important for any number of reasons, from identifying asteroids that might pose a risk to Earth to tracking spacecraft and space debris to avoid collisions. Meanwhile, improved detection could help scientists learn more about the outer reaches of the solar system.
Even as new hardware is developed, such as improved sensors or more advanced telescopes, Nguyen’s algorithmic approach will provide a force multiplier. “You can use it on a ground-based telescope, or with a space-based camera,” she says. “It’s flexible and platform-agnostic. Whatever type of hardware you have available, the algorithm will help you boost the results.”
October 17, 2025
|
