Abstract
In the last decades, there is a growing in the use of UAVs for inspection applications due to their maneuverability, flexibility, and efficiency. UAVs can be used for regular inspections to verify deformities, reconstruct 3D spaces, and mapping through aerial photogrammetry. An important aspect when inspecting buildings and structures is to cover the entire structure as efficiently as possible. Some applications, such as the inspection of electric power generation facilities, demand security constraints to ensure safety in their big structures like slopes and dams. Those inspections will be repeated regularly to ensure that changes in the structures are not occurring over time. The determination of the optimal path planning for those structures can be quite complex. This is due to the presence of obstacles and safety restrictions, such as high voltage power lines, transformers, and water outlets that will be present and can not be modeled previous to the first inspection. Thus, these initial inspections are performed manually by a skilled operator. However, this first path is performed without considering mission time and optimal trajectory. Therefore, this research work proposes a methodology to maximize coverage inspection trajectories generated manually. The results proved the algorithm capacity of optimizing trajectories performed manually by an operator.
