In 2022, Randall Pietersen, a civil engineer in the U.S. Air Force, embarked on a training mission to evaluate damage at an airfield runway, practicing “base recovery” procedures following a simulated attack. For hours, his crew traversed the area in chemical protective attire, communicating geocoordinates as they recorded damage and searched for hazards such as unexploded munitions.
This task is routine for all Air Force engineers prior to deployment, yet it carried particular importance for Pietersen, who has devoted the past five years to creating quicker, safer methodologies for evaluating airfields as a master’s student and presently a PhD candidate and MathWorks Fellow at MIT. For Pietersen, the labor-intensive, meticulous, and potentially perilous assignment highlighted the promise of his research to facilitate remote airfield assessments.
“That experience was truly enlightening,” Pietersen states. “We’ve been informed for nearly a decade that a new, drone-based system is forthcoming, but it remains hampered by its inability to identify unexploded ordnance; from above, they resemble rocks or debris too closely. Even ultra-high-resolution cameras simply do not have adequate performance. Rapid and remote airfield evaluations are still not standard practice. We’re only equipped to undertake this on foot, and that’s where my research fits in.”
Pietersen’s ambition is to develop drone-operated automated systems for evaluating airfield damage and detecting unexploded ordnance. This quest has led him through various research avenues, from deep learning to small unmanned aerial systems to “hyperspectral” imaging, which captures passive electromagnetic radiation across an extensive range of wavelengths. Hyperspectral imaging is becoming more affordable, swifter, and more resilient, potentially rendering Pietersen’s research increasingly relevant across multiple sectors including agriculture, emergency response, mining, and building evaluations.
Discovering computer science and community
Raised in a suburb of Sacramento, California, Pietersen was drawn to mathematics and physics during his schooling. However, he was also a cross-country runner and an Eagle Scout, and he sought a means to combine his passions.
“I appreciated the diverse challenge presented by the Air Force Academy,” Pietersen remarks. “My family lacks a history of military service, but the recruiters spoke about the comprehensive education, incorporating academics along with athletic fitness and leadership. That holistic approach to the college experience resonated with me.”
At the Air Force Academy, Pietersen majored in civil engineering, where he first began to learn how to undertake academic inquiries. This journey necessitated him to acquire a bit of computer programming skills.
“In my final year, the Air Force research laboratories had several pavement-oriented projects that aligned with my role as a civil engineer,” Pietersen recalls. “While my expertise helped outline the initial challenges, it swiftly became evident that crafting suitable solutions would necessitate a profound understanding of computer vision and remote sensing techniques.”
The assignments, which involved airfield pavement evaluations and threat detection, also led Pietersen to begin employing hyperspectral imaging and machine learning, which he furthered upon arriving at MIT for his master’s and PhD in 2020.
“MIT was an obvious choice for my research since the institution has a robust history of research collaborations and interdisciplinary thinking that aid in solving these unconventional issues,” Pietersen states. “There’s no superior place globally than MIT for forefront endeavors like this.”
By the time Pietersen reached MIT, he had also taken up extreme sports such as ultra-marathons, skydiving, and rock climbing. Some of this enthusiasm stemmed from his involvement in infantry skills competitions during his undergraduate years. The multi-day competitions are military-centric races in which teams from various nations navigate mountains and perform graded activities like tactical combat casualty care, orienteering, and sharpshooting.
“The peers I associated with in college were greatly enthusiastic about those activities, so it was a natural progression of relationship-building,” Pietersen notes. “These events would physically challenge you for 48 or 72 hours, sometimes with a little sleep mixed in, and you get to compete alongside your friends and enjoy yourselves.”
Since relocating to MIT with his spouse and two children, Pietersen has connected with the local running community and even served as an indoor skydiving instructor in New Hampshire, although he concedes that the East Coast winters have proven challenging for him and his family to acclimate to.
Pietersen worked remotely from 2022 to 2024, but he wasn’t conducting his research from the convenience of a home office. The training that illuminated the realities of airfield assessments occurred in Florida, after which he was deployed to Saudi Arabia. He even wrote one of his PhD journal articles while stationed in a tent in the desert.
Now back at MIT and approaching the completion of his doctorate this spring, Pietersen expresses gratitude for all the individuals who have supported him along his journey.
“It has been enjoyable to explore various engineering fields, attempting to solve problems with the guidance of all the mentors at MIT and the resources available to address these highly specialized challenges,” Pietersen states.
Research with a mission
In the summer of 2020, Pietersen interned with the HALO Trust, a humanitarian organization dedicated to removing landmines and other explosives from regions affected by conflict. This experience unveiled another significant application for his work at MIT.
“We have post-conflict zones worldwide where children are trying to play, yet landmines and unexploded ordnance are present in their yards,” Pietersen states. “Ukraine serves as a current example in the news today. There are always traces of war left behind. At present, individuals must enter these potentially hazardous areas to clear them, but emerging remote-sensing methods could expedite that process and enhance safety.”
Although Pietersen’s master’s studies primarily concentrated on evaluating standard wear and tear of pavement structures, his PhD has centered on identifying unexploded ordnance and more extensive damage.
“If the runway is attacked, there would be bombs and craters scattered all over it,” Pietersen explains. “This creates a challenging environment for assessment. Various types of sensors gather different kinds of information and each has its advantages and drawbacks. Much work still exists on both the hardware and software fronts, but thus far, hyperspectral data seems to be a promising differentiator for deep learning object detection.”
Upon graduation, Pietersen will be stationed in Guam, where Air Force engineers routinely conduct the same airfield assessment simulations he participated in while in Florida. He aspires that in the near future, those evaluations will be performed not by humans in protective gear, but by drones.
“Currently, we depend on visible lines of sight,” Pietersen comments. “If we can shift to spectral imaging and deep learning solutions, we can finally perform remote evaluations that ensure greater safety for all.”