See-through power: Air Force engineer leverages AI to detect unexploded munitions with drones

Air Force engineer Randall Pietersen is developing innovative drone-based systems to transform dangerous airfield assessments, potentially saving lives and time in military operations. His MIT research combines hyperspectral imaging with machine learning to detect unexploded munitions—a challenge that has stymied previous drone systems which struggle to distinguish ordnance from rocks and debris. This work leverages increasingly affordable and durable hyperspectral technology that captures electromagnetic radiation across broad wavelengths, with applications extending beyond military contexts to agriculture, emergency response, and infrastructure assessment.

The big picture: Pietersen’s research addresses a critical military safety challenge after experiencing firsthand the dangers of conventional airfield assessment during a 2022 training mission.

• Air Force engineers currently conduct post-attack airfield assessments on foot while wearing chemical protection gear, manually documenting damage and identifying unexploded munitions.
• Despite promises of drone-based solutions over the past decade, existing systems remain inadequate due to their inability to reliably identify unexploded ordnance from aerial imagery.
• Pietersen’s personal experience with these dangerous protocols during training exercises has reinforced his commitment to developing safer, more efficient remote assessment technologies.

Key innovation: Hyperspectral imaging combined with deep learning forms the foundation of Pietersen’s approach to remote airfield assessment.

• Unlike conventional cameras that capture visible light, hyperspectral imaging can detect passive electromagnetic radiation across hundreds of wavelengths, revealing subtle material differences invisible to standard sensors.
• The technology is becoming increasingly practical for field deployment as hyperspectral cameras become cheaper, faster, and more durable.
• By using machine learning to analyze this rich spectral data, Pietersen’s system could reliably distinguish between harmless objects and dangerous unexploded munitions from a safe distance.

Broader applications: While developed for military purposes, Pietersen’s research has significant potential across multiple civilian sectors.

• The same technology could benefit agriculture by monitoring crop health, assist in emergency response scenarios, enhance mining operations, and improve building assessments.
• The decreasing cost and increasing durability of hyperspectral imaging systems makes widespread adoption increasingly feasible across these industries.
• These remote sensing capabilities represent a significant advancement in safety protocols for inspection tasks that currently require direct human exposure to potentially hazardous environments.