How can AI help keep our network running during power failures?

The Problem

Extreme weather conditions in Bangladesh can disrupt connectivity by causing a cell site to lose power. Certain areas are more affected than others, but there is a rising trend of more and more extreme weather across the world. The risk of losing grid power to a cell site due to bad weather conditions is mitigated by battery backup systems on the sites that ensures that connectivity stays up - no matter what.

However, the battery capacity is limited, and secondary power sources are needed.

Sometimes, especially during extreme weather, GrameenPhone must manually bring diesel power generators to the cell site to ensure it has power. This means that the process of ensuring constant connectivity is a challenge. The number of these generators is limited, so care must be taken when deciding which site to bring them to. This prioritization task was made harder by the fact that only rough estimates of the battery capacity at a given site was generally available.

The Solution

The Solution
AI Algorithm For Battery Back-up Insight and Optimization

The Telenor Research&Innovation team together with Grameenphone designed an AI algorithm that can predict remaining battery capacity on the site level.

The algorithm predicts remaining uptime from the battery when power goes out. This enables the engineers to better plan how mobile power solutions should be rolled out during extreme weather events.

The Results

The Results
Reducing carbon emissions and cost of operation

This innovative application of the AI model allows engineers to prioritize sites for battery replacements with diesel generators.

The Value Proposition

Increased efficiency of choice

The AI can support the on-site agents in choosing where to roll out portable generators more efficiently, taking a step towards minimising the communication consequences from challenging weather conditions.