Rapid Automated Damage Detection with Autonomous Robots (RAD²AR)

Background

Natural disasters, such as earthquakes, hurricanes, and flooding, pose a significant threat to the stability and strength of our infrastructure. Every year, the U.S. invests $12 billion on bridge repairs, while 60,000 bridges are rated as “structurally deficient.” In the case of a natural disaster, such infrastructure will likely become further damaged, posing a threat to public safety. The project focuses on the Seattle-metro area, where current infrastructure is threatened by earthquakes. The project investigates the use of small unmanned aerial systems to quickly assess the damage of a structure following a natural disaster. Research being conducted by Pacific Northwest Economic Region (PNWER) and faculty at Northeastern University have developed a method use sUASs to model, analyze, and assess any damaged structures.

Goals and Objectives

The objectives and expected outcomes of the project include:

• Develop the technology to use small unmanned aerial systems to autonomously circumnavigate a structure and automatically assess the damage on the structure following a natural disaster based on using 3D laser point clouds and images:

    • Develop algorithms for combining registered 3D point cloud maps and images to reconstruct the geometry of structures.

    • Use heuristic techniques and machine learning to automatically identify and document damaged areas of concern

• Communicate findings to emergency response managers who can determine the most efficient and effective ways to address the damaged areas.