Electrical substations are crucial for the power grid, often situated in remote areas, posing logistical challenges for maintenance. Unexpected faults can lead to network failures or accidents. Hence, a methodology employing non-invasive, remote monitoring is proposed to predict and diagnose issues like partial discharges or heating.

This approach involves low-cost sensors and a communication network to gather data on substation status. Automated analysis using AI methods creates indicators for current and future equipment status. The system includes a data concentrator, sensors for various parameters, and software for configuration and alerts.

Installation is straightforward, and components require minimal maintenance. The system reduces inspection needs, minimizes downtime, and enhances safety by enabling real-time monitoring. It also helps in decision-making for corrective, preventive, or predictive maintenance. Additionally, variations in quantities can explain dynamic events, such as weather impacts on substation components.

Overall, the system aims to mitigate accidents, reduce technical losses, ensure service continuity, and contribute to technological advancements in the electricity sector.