The challenge
To improve efforts to identify property features such as buildings, Enel-Codensa faced the challenge of improving the accuracy and timeliness of early detection of encroachments on power transmission lines. In addition, its extensive network coverage required a robust technology solution to automate its classification initiative.
The solution
To address the problem of manually evaluating images to detect buildings in areas spatially adjacent to high-voltage power lines in different geographic regions, Servinformacion used Google Cloud Platform to train a convolutional network that classifies, locates, segments, and then returns the results for visualization and decision-making.
Project results
By incorporating Machine Learning technology through a segmentation model using Google Cloud's AI/ML capabilities, Enel-Codensa immediately improved its easement management process, achieving a high level of reliability and complete visualization of detected features in BigQuery.
Servinformación helped identify the opportunity to improve early detection of encroachments on state-owned strips of land, enabling employees to quickly assess the full extent of their network coverage. This process had previously been performed manually without achieving 100% coverage.
About the Customer
Enel–Codensa is a leading electricity distribution and retail company in the Colombian market, with more than 3.4 million customers across Bogotá, Cundinamarca, eight municipalities in Boyacá, one in Tolima, one in Caldas, and one in Meta.
Industry: Energy Location: Colombia
