TETRA Autodrone: Automated inspection of infrastructure using drones


Contact: Steve Vanlanduit

Background

Worldwide there is a strong growing demand for efficient inspection of infrastructure (including bridges, locks, buildings, wind turbines, solar panels, etc.). Recent technological developments around drones equipped with cameras can facilitate these inspections.

Although there are now about 500 certified drone pilots active in Belgian drone service companies, the number of drone inspections carried out to date has been very limited. One of the reasons for this underutilized potential is that there is insufficient knowledge within drone service companies about image capture and image processing to be able to perform these inspections in a systematic and high-quality manner.

The research groups of the University of Antwerp and the WTCB do have this expertise and will share it with the above-mentioned target groups in this project.

In the project, nine types of drone inspections are performed by drone service companies from the user group. These demonstrators include inspections ranging from roads to structures, masts, buildings and perhaps even offshore installations.

On the basis of these case studies, an in-depth investigation will be carried out into the needs and requirements of these parties in order to be able to carry out more efficient inspections and to subsequently tailor the tools to be developed accordingly. These tools are subsequently extensively tested on the basis of these cases.

A demonstration video and a report are produced for each of the nine drone inspection case studies. These are made available on a project website and distributed via social media (YouTube channel). Furthermore, a broader target group of companies is reached through newsletters, events and websites of the collective centers that are involved in the user group (Confederatie Bouw, EUKA, IBN Offshore and VITO).

In addition, co-applicant WTCB is also jointly responsible for the dissemination of the project results through their dissemination channels (website, publications, events, personal contacts). The bachelor's and master's students of the Industrial Engineering, Electromechanics and Architecture courses of the University of Antwerp are actively involved in the implementation of the project (at least eight master's theses and also the integration of the project results in at least ten course units).


Aim of the project

The overall aim of the project is to develop, validate and demonstrate a methodology for the efficient detection and monitoring of damage to infrastructure using drones. The project is primarily aimed at increasing knowledge about image capture and image processing among the 500 certified drone pilots currently in Belgium. By leveraging the project results, they can offer reliable automated and repeated drone inspections.

In concrete terms, in this project, based on prior research by UAntwerp and WTCB, we are realizing an open source software toolchain for automated damage detection and monitoring of complex 3D structures.

By combining and expanding existing tools and documenting the software with real-life cases, we ensure that drone service companies with limited programming experience can offer inspections in a more quality, reliable, repeatable and time and cost efficient way.