The project proposes an autonomous visual inspection system which will enable inspection experts, untrained in piloting micro aerial vehicles (MAVs), to tele‐operate them as an aid to their mission, while focusing on the inspection task at hand. The inspection task we propose to solve is visual inspection of wind turbines while in operation, a highly complex task which calls for new theoretical and practical paradigms. Existing inspection methods are slow and costly, and the wind turbine must be out of service and stopped while inspected, leading to unavailability and causing significant loss of electricity production and income. To successfully solve this task, it is necessary to bring together different expertise emanating from research and industry. Accordingly, this project proposal is the joint effort of the University of Zagreb, Faculty of Electrical Engineering (UNIZG-FER), University of Dubrovnik (UNIDU) and DiagnostiQA Consultoria Tecnica (DQA). Together we participate in the EUROC Challenge 3 under the name BladeHunters (more information about us at https://sites.google.com/site/eurocunizgfer/home). The proposed system will be developed in 26 months through Stage II – Realistic Labs and Stage III – Field tests of the EUROC Challenge. In Stage IIa – Benchmarking + Free-style round, we will show our capability to solve a use-case driven task defined by the challenge host and a task defined by our team chosen to demonstrate our know-how not entirely included in the first task. Successful completion of this round shows our skill to tackle new problems and solve them in the Realistic Lab setting, thus enabling us to proceed to the next round Stage IIb – Showcase, the end-user driven task. Bringing a real-life scenario, proposed in Stage III of the project, to the Realistic Lab setting within this round, will enable us to increase the Technology Readiness Level of the proposed aerial system, thus shorten the time needed for Field-test implementation at Stage III. Development of a comprehensive aerial system for autonomous inspection of rotating turbine blades in real operating conditions requires the expertise of all team participants whose background and previous experience corroborate their commitment to this project and the capabilities in delivering the tasks assigned to them. Altogether, participants collectively constitute a team capable of achieving the project objectives, broken down into tasks and assigned to participants based on their fields of expertise. This project will establish new systematic foundations within methods and computational schemes for MAVs navigation and control. The project participant teams are scientific and technological leaders in their respective areas and therefore the project will contribute to maintaining this position by renewed methodological development aimed at addressing problems derived from complex task of wind turbine blades inspection during normal operation. While DQA as the end user of the new inspection system will gain significant advantage with respect to the competitors on the market of wind farm inspections in Europe and worldwide, UNIZG-FER and UNIDU as the research partners will benefit from experimental testing of their algorithms for MAV navigation and control in the very complex real-life scenario.
Duration: July 2015 – October 2016