UVM Reference Number: C0961i
Summary:
Summary:
MARSBots are low-cost swarm microrobots equipped with cameras and sensors that use augmented reality (AR) for the autonomous inspection of hard-to-reach structural areas, detecting cracks and early signs of damage.
Technology Overview:
MARSBots are small, flexible-legged microrobots designed to inspect confined or hazardous spaces without human entry. Each robot uses an ESP32-Cam for image capture and transmits data wirelessly to an edge processor. Operators can control the swarm through AR interfaces on a HoloLens 2 headset or PC, which display sensor data and visual feedback in real time. The system integrates AI-powered semantic segmentation to identify cracks, using machine learning models trained and fine-tuned on diverse datasets. MARSBots make structural health monitoring faster, safer, and more precise than traditional manual inspections or single-robot systems by combining swarm mobility, multi-view imaging, and AR visualization.
Key Advantages:
- Enables inspection of confined or dangerous areas without human entry
- Uses cognitive swarm behavior for full-area coverage
- Low-cost, lightweight, and easy to produce with 3D-printed parts
Problems Solved:
- Reduces risk to human inspectors in hazardous environments
- Prevents structural collapses through early detection of damage
- Overcomes limitations of manual or single-robot inspections
Applications:
- Structural health monitoring in bridges, tunnels, and buildings
- Post-disaster infrastructure inspection
- Industrial facility maintenance
- Civil engineering research and safety assessments
Seeking
