I have been working with drones since 2014, however the outbreak of war in Ukraine marked a turning point in my profession. Since 2022, my focus has shifted to exploring how drones will be used to automate humanitarian demining – what capabilities they need, and the way technology could make these efforts safer and more efficient. As a part of this work, I closely follow the Geneva International Centre for Humanitarian Demining (GICHD), attend their events, and repeatedly engage with their experts.
Considering drone-based solutions paired with AI, they are literally helpful only on the non-technical survey (NTS) stage of the humanitarian demining process. It means drones scan large areas and collect data. Then, a machine learning model analyzes this data to flag regions that contain mines. Not the precise places of mines.
Technical survey (TS), which confirms and maps contaminated areas, still relies on personnel with metal detectors, trained dogs, and mechanical demining machines. They go into the mined area to pinpoint the precise locations of the hazards.
The method keeps being long, dangerous, and expensive:
Mines also proceed to be a threat to civilians – there have been at the very least 5,757 mines/ERW casualties in 2023.
On this post, I explain why current drone-based solutions don’t work for technical survey (the costliest and time-consuming stage at once) and share what I see as the perfect technique to fix that.
Detecting mines under soil or vegetation is sort of unattainable
Drones with standard optical or thermal cameras often capture images from a single downward-facing angle. This approach works well for spotting surface-level anomalies but fails to detect buried or hidden mines. For that reason, drones are mostly used for non-technical surveys in humanitarian demining.
One in all the frontline solutions – Secure Pro AI – reports that they’ve only a 5 percent detection rate in regions with trees and bushes.
Regardless that it’s less relevant to Ukraine, where most mines are scattered on the bottom, as a substitute of buried, the situation could be very different (for instance) for Cambodia:
- 4-6 million landmines remain from conflicts within the Seventies-90s
- 64,000+ casualties since 1979, with children as primary victims
Non-metal and old metal mines are harder to detect, even on the surface
Non-metal mines present a good portion of landmines in current and former conflict zones. They’re intentionally designed to bypass detection by conventional metal detectors.
Visually, non-metallic mines are hard to detect. They don’t shine, stand out in images, or show up well on thermal cameras. Metal detectors and magnetometers either miss them or trigger too many false alarms.
So, current drone-based detection tools often miss non-metallic mines entirely.
In terms of old metal mines, corrosion changes how they appear and behave, in order that they mix into the bottom and respond poorly to detection tools. Misshapen ones are even harder to discover in images.
And since these mines are harder to identify, they take for much longer to seek out and take away, or they stay hidden and put each deminers and civilians in danger.
Weather and daytime dependency
If we’re talking about drones with RGB and multispectral cameras, they require daylight. In cloudy, low-light, or shaded areas (forests, ruins), image quality and object detection drop too.
Thermal detection, in its turn, works best at dawn or dusk, when the bottom and mine differ in temperature. During midday, the sun heats every part equally, reducing contrast.
While rain and wet soil blur surface detail, alter soil color and temperature, and might hide soil disturbance or thermal anomalies. Snow just covers visual markers and equalizes surface temperature, making mines undetectable.
Flying drones only at certain times considerably slows down even the NTS stage of demining, especially in areas with unpredictable weather.
The technology could be very expensive
In 7 affected countries estimated antipersonnel mine contamination area reaches over 100km².
Based on tests in Ukraine, demining with the brand new tech can costs from $3000-5000 to $600-800 per hectare, which remains to be $70,000 per square kilometer. And in some areas, it might exceed the land price itself.
The important reason for the high costs is the multiple false alarms treated as real threats. On average, a team clears over 50 suspected mines to seek out only one actual landmine.
Most heavily contaminated areas are in developing countries. They cannot afford demining without funding from international organizations or governments.
The prices are also too high for businesses to leap in. Once demining becomes low-cost enough, firms might lease mine-contaminated land on the condition that they clear it. In return, they’d get long-term use for a symbolic price and a few tax breaks.
An answer?
With my team, we explored methods that gather more data, can see through foliage and soil, and still maintain sufficient resolution.
An example of a promising development direction is a project by researchers on the University of Oviedo. They’re testing an array-based ground-penetrating synthetic aperture radar (GPR-SAR) system mounted on a UAV.
Their in-flight validation in realistic scenarios proved that the technology solves the next problems:
1) The radar pinpoints the mine’s location with precision, leaving only the disarming or destruction to be done manually.
With the usage of all possible radar paths (fully multistatic configuration), they got high-resolution images where buried targets appeared brighter and clearer. And were in a position to detect with precision difficult targets resembling small, nonmetallic, and shallowly buried objects like plastic anti-personnel landmines, picket pressure plates, and PVC pipes.
2) The answer can work day or night, in varied weather, and even with moderate vegetation.
How it really works:
- Sends radar pulses into the bottom.
- Detects reflections from subsurface changes (e.g., plastic, metal, voids).
- Builds 3D subsurface images with centimeter-level accuracy by combining radar signals from multiple transmitter-receiver (Tx- Rx) pairs and flight positions.
The answer still has its limitations, but based on my background, it’s probably the most relevant direction of research and development at once.
One in all GPR’s important strengths is how much data it will possibly collect. More data means researchers can improve accuracy at the popularity/classification stage with AI. This results in more efficient survey and clearance work and cuts overall costs by 50% or more.
