Drones and AI Accelerate Land Mine Removal

Researchers develop new techniques to improve speed, accuracy and reliability of detecting buried explosives.

Published on Mar. 6, 2026

Got story updates? Submit your updates here. ›

Researchers are using drones equipped with multiple sensors and AI algorithms to speed up the detection and removal of land mines, which continue to kill and injure thousands of civilians each year in conflict zones around the world. The team has created new benchmark datasets to help advance this technology, which could transform land mine detection from a slow and dangerous manual process into a safer, smarter and more scalable operation.

Why it matters

Land mines remain a major threat to civilian populations in many post-conflict regions, killing and injuring thousands each year. Traditional ground-based detection methods are slow, resource-intensive and put deminers at risk. Drone-based aerial surveys combined with AI-powered data analysis could dramatically improve the speed, accuracy and safety of land mine removal efforts.

The details

The researchers are using a combination of sensors on drones, including RGB cameras, thermal imagers, multispectral and hyperspectral sensors, LiDAR, and electromagnetic detectors, to gather comprehensive data on suspected minefields. They have also created large, publicly available benchmark datasets that include precise ground truth on the location of inert mines, which will help advance AI-powered detection algorithms. A key focus is developing methods to estimate the uncertainty of AI predictions, to provide deminers with a clearer sense of confidence in the detection results, especially in challenging conditions.

  • In 2024 alone, 1,945 people were killed by mines and 4,325 were injured, 90% of whom were civilians.
  • Demining operations removed 105,640 mines in 2024.

The players

Sagar Lekhak

A Ph.D. student in the Imaging Science Department at Rochester Institute of Technology, working on using drone-based, multisensor imagery and artificial intelligence to improve land mine and unexploded ordnance detection.

Emmett Ientilucci

A professor at Rochester Institute of Technology, collaborating with Sagar Lekhak on this research.

Demining Research Community

A nonprofit organization that provided a controlled test field in Oklahoma for the researchers to collect comprehensive multisensor datasets on land mine and unexploded ordnance targets.

Royal Military Academy of Belgium

Collaborated with the researchers on a large data collection campaign, deploying over 110 replicas of PFM-1 mines across varied terrains and vegetation conditions.

Got photos? Submit your photos here. ›

What they’re saying

“At its core, this research is not about algorithms or drones, it is about people. It is about farmers reclaiming their land, children walking safely to school, and communities rebuilding without fear.”

— Sagar Lekhak, Ph.D. Student in Imaging Science, Rochester Institute of Technology

What’s next

The researchers plan to release the full multisensor dataset they collected through an upcoming journal publication, which will provide new opportunities for the broader AI and remote sensing community to advance land mine detection technology.

The takeaway

By combining drone-based multisensor data collection and AI-powered detection algorithms, researchers are working to transform land mine removal from a slow, dangerous manual process into a safer, smarter and more scalable operation that can help restore post-conflict regions and save lives.