VEGAnalytics, a project developed by EDP Distribuição with the support of EDP Inovação, is a tool for controlling the protection zones around high and medium voltage overhead lines.

The tool calculates vegetation growth using various parameters and variables to predict the growth of each plant species, using artificial intelligence techniques, resulting in the generation of alerts of potential vegetation interference situations with the power grid. If the landowners have not acted, in view of a risk situation of people and property, work orders are taken to prune or cut the vegetation.

In this way, EDP Distribuição is able to monitor the 68,000 km of AT and MT power lines against regulated safety distances, and replace owners when the safety of people and property is compromised.

This project uses the LIDAR system, a laser to measure and map 3D information about the environment around power lines.
 

Thus, the VEGAnalytics project goes through several steps:

• 1) Map the power lines: LIDAR is installed on a helicopter or unmanned aerial vehicle (UAV). The vehicle flies over the power lines to map them. This technology emits a laser pulse so that reflected light is collected in the receiver. The LIDAR can reach 200,000 pulses per second. The result of this process is accurate 3D mapping along the power line, where we can see not only soil, vegetation, trees and artificial objects, including poles and lines, but also noise (eg birds).
• 2) Advanced Analytics for Power Lines Inspection Algorithm: Since the data collected contains information not relevant for analyzing the distance from vegetation to power lines, LABELEC has developed an algorithm for processing LIDAR data. Thus, this technology generates reports of data captured on flights, such as the type of obstacle identified, distance to line, among other information.
• 3) Vegetation growth prediction algorithm: Reports generated by LABELEC are loaded into a geographic database and under this information an algorithm developed by EDP Distribution of tree growth prediction along lines, based on several factors, is applied. as the growth rates of the trees, the soil type of the place, the climate type, among others From these factors, it is then possible for the developed algorithm to estimate how far a particular tree from the line will be in one year, two years, or three years if it is not cut or trimmed. From these growth results, files are generated which indicate, taking into account the stipulated safety distances, which trees potentially need intervention, grouped by span.
• 4) Reporting of results to the field: Once the LABELEC reports have been processed and the results obtained by the algorithm indicating the trees that potentially need intervention are obtained, they are sent to EDP Distribuição's providers to carry out the work.