**4. Discussion and conclusion**

The inspection of tunnels and infrastructure for water and hydroelectric resources and, more generally, for any underground work, is essential for the efficient maintenance of the infrastructure itself and provides significant benefits for the rational use of resources. This type of activity collects important information about the current state of consistency of the structure. By highlighting potential or actual failure conditions such as cracks, deformations or other types of problems, it is possible to plan any safety maintenance operations in a timely way. These inspections are now carried out mainly by human operators, with considerable risks to their safety and health at work: claustrophobic, dark and dirty environment.

The idea of this project is to apply innovative techniques, to overcome these problems with the future purpose to ensure greater safety, avoid the inconvenience and risks arising from these activities for the human operator and meet the market needs. As a consequence, a scheduling system has been presented and allows to set different strategies to approach the inspection of the tunnel before starting the mission. Autonomous driving techniques in the six degrees of freedom are developed to ensure the obstacle avoidance in confined space using a simple Lidar sensor. By applying visual-inertial odometry and its fusion with the aid of a Kalman filter, it has been possible the realization of a UAV system able to perform an autonomous inspection of indoor environment like tunnels or conduits.

Although the results shown in this work in terms of robustness and consistency are encouraging, in the future there will be a need to develop advanced techniques considering different scenarios and environments. One possible improvement could be brought developing navigation algorithm based on other types of sensor and using alternative approach. In conclusion, it is central to continue to investigate visual-inertial algorithm since its contribution has proven essential for the robustness, reliability and efficiency of the overall system.
