A helicopter flying just 100 feet above the ground may have only seconds to detect radioactive material hidden among the natural radiation emitted by the surrounding environment. Identifying the material is difficult enough. Determining its exact location and how radioactive it is, without stopping to take additional measurements, presents an even greater challenge.
Now, researchers in Switzerland have demonstrated a new method that can accomplish all three tasks using just six seconds of radiation measurements collected during a single helicopter pass.
In field experiments, the system correctly distinguished between different radioactive materials, located their positions to within approximately six feet, and estimated their radioactivity with errors as low as 2%.
According to the researchers, the breakthrough addresses a longstanding limitation in mobile radiation detection.
The findings, published in Nature Communications, may greatly improve how emergency responders investigate nuclear accidents, militaries search for radioactive threats, and scientists map radioactive elements on Earth and other planetary bodies.
“The results presented here show that all three quantities can be recovered jointly within a single full-spectrum Bayesian framework, with percent-level activity accuracy, meter-scale localization accuracy, and decisive source discrimination demonstrated under realistic single-pass survey conditions,” researchers write. “This approach provides the methodological and computational foundation for a new generation of quantitative mobile gamma-ray spectrometry.”
Finding Radioactive Material While Moving
Gamma rays are high-energy electromagnetic radiation emitted by certain radioactive materials. Specialized detectors can measure their energy footprints, helping scientists identify the elements responsible.
However, identifying radioactive sources from a moving aircraft is considerably more complicated than analyzing them in a laboratory.
As a helicopter moves, the distance and angle between its detector and a radioactive source constantly change. The aircraft’s structure can also absorb or scatter radiation, distorting the measurements.
Naturally occurring radioactive elements in the ground, including potassium, uranium, and thorium, create background signals that can also obscure the materials investigators are attempting to identify.
Traditional approaches often struggle to separate overlapping signals, particularly when detectors have only a few seconds to collect data.
Researchers at Switzerland’s Paul Scherrer Institute, working with the University of Michigan, developed a computational system designed to tackle these limitations.
Their technique combines detailed computer simulations of how radiation travels through the environment and aircraft with Bayesian statistical inference, which evaluates competing explanations for the detector’s measurements.
Rather than relying on individual peaks in the radiation spectrum, the method evaluates the entire signal, including information that typical approaches could miss.
Six Seconds of Data Reveal Two Radioactive Sources
To test the technique, researchers analyzed measurements collected during a June 2022 exercise at a military training ground near Thun, Switzerland.
A Super Puma helicopter equipped with a gamma-ray spectrometer flew over sealed radioactive sources containing cesium-137 and barium-133.
The experiments included stationary measurements from approximately 295 feet above the ground and moving surveys conducted at roughly 100 feet.
During the moving tests, the helicopter traveled at approximately 22 miles per hour, collecting radiation measurements once every second.
Researchers examined three configurations: cesium-137 alone, barium-133 alone, and both materials positioned less than eight inches apart.
For each configuration, the system analyzed a six-second measurement window. In all three scenarios, the method correctly identified which radioactive materials were present, with statistical evidence exceeding the five-sigma threshold commonly associated with highly significant scientific findings. It also located the sources to within approximately 6.6 feet, comparable to the helicopter’s satellite-positioning uncertainty.
For the individual radioactive sources, estimated activity levels differed from independently measured laboratory values by less than 2%. Even when both sources were present, the estimates remained within 4%.
Perhaps the most revealing result came from the combined-source experiment. The researchers clearly identified the weaker barium-133 signal despite its being substantially obscured by the stronger cesium-137 radiation.
The ability to distinguish between different radioactive materials is particularly important in real-life situations, where multiple sources may produce overlapping radiation signals that make it difficult for investigators to determine exactly which substances are present.
A Major Computational Obstacle Overcome
One of the study’s more consequential advances involves how the researchers handled the enormous computing requirements of correctly modeling radiation.
In-depth simulations must factor in the helicopter’s structure, changing position, fuel load, and surrounding atmospheric conditions. Running these simulations repeatedly during statistical analysis would ordinarily require unreasonable computing resources.
Instead, researchers performed the most demanding radiation-transport calculations beforehand, creating reusable libraries that could be assembled into new predictions. This reduced the computational cost of individual high-fidelity model evaluations by approximately 10 million times, according to the study.
However, that improvement should not be confused with a complete radiation-identification system operating instantaneously. The researchers demonstrated accurate analysis of short measurement windows, not a fully deployed real-time emergency-response capability.
From Nuclear Accidents to Other Worlds
The breakthrough’s implications extend well beyond military and national security efforts to detect and identify potentially dangerous radioactive materials. Following a nuclear accident, airborne radiation surveys can help authorities identify contamination and determine where more detailed ground investigations are needed.
The same analytical approach could eventually support robotic vehicles and drones operating in hazardous environments where sending personnel would be perilous. Researchers also see applications in planet exploration, where spacecraft employ gamma-ray measurements to investigate the elemental composition of planetary surfaces.
The technique could help extract more reliable information from limited observations, possibly boosting measurements of radioactive elements on the Moon, Mars, and other planetary bodies.
That said, important challenges remain. The experiments involved known radioactive materials under controlled field conditions, rather than an unknown contamination event. Extending the method toward widespread contamination, more complicated environments, and operational deployments will require additional development.
Nevertheless, the findings show that even a few seconds of radiation data can contain substantially more useful information than conventional analysis methods can reliably recover.
For investigators searching for radioactive material, that could eventually mean the difference between identifying a suspicious signal and knowing exactly what produced it, where it originated, and how radioactive the source actually is.
The recent study, “Quantitative mobile gamma-ray spectrometry through Bayesian inference,” appeared in Nature Communications.
Tim McMillan is a retired law enforcement executive, investigative reporter and co-founder of The Debrief. His writing typically focuses on defense, national security, the Intelligence Community and topics related to psychology. You can follow Tim on Twitter: @LtTimMcMillan. Tim can be reached by email: tim@thedebrief.org or through encrypted email: LtTimMcMillan@protonmail.com