Unmanned Aerial Vehicle Remote Sensing Technology for Structural Damage Assessments in Low-Light Conditions
Abstract
The research explores the viability of using a small Unmanned Aerial Vehicle equipped with thermal imaging and lowlight camera to assess structural damage to steel girders. Damage assessments following natural disasters are daunting and arduous tasks that are resources intensive and dangerous. Unmanned aerial vehicles with remote sensing technology (UAV-RS) have been used in recent large-scale disaster events such as Hurricanes Katerina, Harvey, Irma, and Maria as well as others. Current assessment methods of structures include; inspectors physically conducting detailed and rapid surveys of damage with or without the assistance of special equipment, use of helicopters, satellite imagery, and new innovative methods using unmanned aerial vehicles with remote sensing technology. The initial experiment utilized the S-BRITE facility at Purdue University. Two steel girders located at S-BRITE were used in the experiment with damages that render them structurally deficient. Experiments were conducted during hours of low visibility. Most scientific studies have focused on using UAV-RS during hours of daylight. This research explores the use of UAV-RS during low-light conditions (i.e. early evening nautical and astronomical twilight, and night) for detecting global damage to steel girders. The goal is to present evidence for further study in the use of UAV-RS during low-light conditions for inspecting structures to include primary load bearing members. The research concluded that while the UAVRS can detect global damage in low visibility conditions, further experiments in varying low-light conditions to include 3D imaging and semi-autonomous inspection using computer vision are important for structural damage assessments.
Degree
Ph.D.
Advisors
Rapp, Purdue University.
Subject Area
Aerospace engineering|Political science|Public administration|Remote sensing|Robotics|Transportation
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.