The problem
Solvane’s inverters report thousands of readings a minute. Almost all of it is normal. The few readings that matter were buried, and crews were dispatched on a fixed schedule instead of on evidence.
What changed
Kairo learned the normal output curve for each site, accounting for weather, season and panel age. When a string started drifting, it flagged it with the likely cause and the cost of waiting.
- Daily dispatch list ranked by output at risk
- Weather-adjusted baselines for every site
- Briefs sent straight to the regional crew lead
The outcome
Faults are found a week earlier on average, and almost a third fewer trucks leave the depot without a clear job to do.
