In 2022, Randall Pietersen, a civil engineer within the U.S. Air Drive, set out on a coaching mission to evaluate injury at an airfield runway, working towards “base restoration” protocol after a simulated assault. For hours, his crew walked over the realm in chemical safety gear, radioing in geocoordinates as they documented injury and seemed for threats like unexploded munitions.
The work is customary for all Air Drive engineers earlier than they deploy, nevertheless it held particular significance for Pietersen, who has spent the final 5 years creating sooner, safer approaches for assessing airfields as a grasp’s pupil and now a PhD candidate and MathWorks Fellow at MIT. For Pietersen, the time-intensive, painstaking, and probably harmful work underscored the potential for his analysis to allow distant airfield assessments.
“That have was actually eye-opening,” Pietersen says. “We’ve been instructed for nearly a decade {that a} new, drone-based system is within the works, however it’s nonetheless restricted by an lack of ability to determine unexploded ordnances; from the air, they give the impression of being an excessive amount of like rocks or particles. Even ultra-high-resolution cameras simply don’t carry out nicely sufficient. Fast and distant airfield evaluation is just not the usual observe but. We’re nonetheless solely ready to do that on foot, and that’s the place my analysis is available in.”
Pietersen’s purpose is to create drone-based automated programs for assessing airfield injury and detecting unexploded munitions. This has taken him down numerous analysis paths, from deep studying to small uncrewed aerial programs to “hyperspectral” imaging, which captures passive electromagnetic radiation throughout a broad spectrum of wavelengths. Hyperspectral imaging is getting cheaper, sooner, and extra sturdy, which might make Pietersen’s analysis more and more helpful in a spread of purposes together with agriculture, emergency response, mining, and constructing assessments.
Discovering pc science and neighborhood
Rising up in a suburb of Sacramento, California, Pietersen gravitated towards math and physics at school. However he was additionally a cross nation athlete and an Eagle Scout, and he wished a solution to put his pursuits collectively.
“I favored the multifaceted problem the Air Drive Academy introduced,” Pietersen says. “My household doesn’t have a historical past of serving, however the recruiters talked in regards to the holistic schooling, the place lecturers have been one half, however so was athletic health and management. That well-rounded method to the school expertise appealed to me.”
Pietersen majored in civil engineering as an undergrad on the Air Drive Academy, the place he first started studying the right way to conduct tutorial analysis. This required him to be taught somewhat little bit of pc programming.
“In my senior 12 months, the Air Drive analysis labs had some pavement-related initiatives that fell into my scope as a civil engineer,” Pietersen remembers. “Whereas my area data helped outline the preliminary issues, it was very clear that creating the best options would require a deeper understanding of pc imaginative and prescient and distant sensing.”
The initiatives, which handled airfield pavement assessments and risk detection, additionally led Pietersen to start out utilizing hyperspectral imaging and machine studying, which he constructed on when he got here to MIT to pursue his grasp’s and PhD in 2020.
“MIT was a transparent selection for my analysis as a result of the varsity has such a robust historical past of analysis partnerships and multidisciplinary pondering that helps you resolve these unconventional issues,” Pietersen says. “There’s no higher place on the earth than MIT for cutting-edge work like this.”
By the point Pietersen acquired to MIT, he’d additionally embraced excessive sports activities like ultra-marathons, skydiving, and mountaineering. A few of that stemmed from his participation in infantry expertise competitions as an undergrad. The multiday competitions are military-focused races during which groups from all over the world traverse mountains and carry out graded actions like tactical fight casualty care, orienteering, and marksmanship.
“The group I ran with in school was actually into that stuff, so it was type of a pure consequence of relationship-building,” Pietersen says. “These occasions would run you round for 48 or 72 hours, generally with some sleep combined in, and also you get to compete along with your buddies and have an excellent time.”
Since coming to MIT together with his spouse and two kids, Pietersen has embraced the native operating neighborhood and even labored as an indoor skydiving teacher in New Hampshire, although he admits the East Coast winters have been powerful for him and his household to regulate to.
Pietersen went distant between 2022 to 2024, however he wasn’t doing his analysis from the consolation of a house workplace. The coaching that confirmed him the truth of airfield assessments came about in Florida, after which he was deployed to Saudi Arabia. He occurred to write down one in every of his PhD journal publications from a tent within the desert.
Now again at MIT and nearing the completion of his doctorate this spring, Pietersen is grateful for all of the individuals who have supported him in all through his journey.
“It has been enjoyable exploring all types of various engineering disciplines, making an attempt to determine issues out with the assistance of all of the mentors at MIT and the sources obtainable to work on these actually area of interest issues,” Pietersen says.
Analysis with a objective
In the summertime of 2020, Pietersen did an internship with the HALO Belief, a humanitarian group working to clear landmines and different explosives from areas impacted by warfare. The expertise demonstrated one other highly effective software for his work at MIT.
“We’ve post-conflict areas all over the world the place youngsters try to play and there are landmines and unexploded ordnances of their backyards,” Pietersen says. “Ukraine is an efficient instance of this within the information in the present day. There are all the time remnants of warfare left behind. Proper now, individuals have to enter these probably harmful areas and clear them, however new remote-sensing methods might velocity that course of up and make it far safer.”
Though Pietersen’s grasp’s work primarily revolved round assessing regular put on and tear of pavement buildings, his PhD has centered on methods to detect unexploded ordnances and extra extreme injury.
“If the runway is attacked, there could be bombs and craters throughout it,” Pietersen says. “This makes for a difficult atmosphere to evaluate. Various kinds of sensors extract totally different varieties of knowledge and every has its professionals and cons. There may be nonetheless a number of work to be accomplished on each the {hardware} and software program aspect of issues, however thus far, hyperspectral information seems to be a promising discriminator for deep studying object detectors.”
After commencement, Pietersen might be stationed in Guam, the place Air Drive engineers often carry out the identical airfield evaluation simulations he participated in in Florida. He hopes sometime quickly, these assessments might be accomplished not by people in protecting gear, however by drones.
“Proper now, we depend on seen traces of website,” Pietersen says. “If we will transfer to spectral imaging and deep-learning options, we will lastly conduct distant assessments that make everybody safer.”
