HYDRO ENVIRONMENTAL AND ELECTRICAL ASSETS MAPPING TOOL USING SWARMS AND MACHINE LEARNING, UAV & USV

Abstract

This paper describe the development of aerial and aquatic drones, using the swarm and machine learning concept to do the management of these equipment, in order to working together in collaborative and autonomous way to collect data, learn the electrical assets, environmental patterns and mapping our dams and power lines, in shortest time and with more mission frequency compared with the current methodology. In Brazil, the dams are used to supply water and generate electricity to the population. The monitoring of water resources is very important to check the degradation indices of the areas and maintain the health of rivers and reservoirs. The power lines electrical assets monitoring process, is very important to keep the maintenance process of their lines updated, to provide energy from the Dams to the Cities. In the actual days, the current monitoring process has been proved inefficient, due to the lack of technical and financial resources in face of the size of the areas to be monitored. This paper reflect the actual stage of our development, it is an improvement from the last paper, presented at IEEE Aerospace Conference 2019, were now is considered the use of the aerial drones with Artificial Intelligence for electrical assets monitoring process. The same considerations of environmental changes are considered here, the robotics systems all the time needed to update themselves according to the dynamic field to perform the mission planned. The Cities, Dams and rivers are constantly changing due to the human or natural actions, like buildings, storms, winds and etc. For this reason, the operational scenery is all the time changing, which request tools to give certain intelligence to the robotics equipment's allowing the machines to “observe the environmental” and “learn” according to these changes. Our proposal is an alternative tool, using aerial (UAV) and aquatic (USV) drones, working together to get information about field, in terms of obstacles, the changes in the field between the monitoring process (missions), learn these changes and give certain intelligence to the drones to make changes automatically in their navigation routes under the mapping/monitoring process. Due the Cities, the Dams size, and their location, in general close to the cities, and considering the lack of resources, the use of swarm concept applied to the aerial and aquatic drones is extremely necessary. The mission places will be divided by zones, each zone will be defined according to the drone operational envelop (e.g., the operational range (time and distance) and terrain geography), more than one aerial drone or USV system can be used to mapping the areas, in this case the swarm concept is applied, at squadron setup, to do the management of the drones. The use of both systems together, for each selected place, will increase the mapping effectiveness, allowing the mapping process of Dams or/and electrical assets in a shortest time, compared to the current methodology, resulting in a complete map of these regions, helping at preventive maintenance process of the Brazilian Energetic Matrix.