SolarMindThe short answer
AI battery optimisation uses information about future prices, solar generation and household demand to help decide when a battery should charge, discharge or retain energy. The label alone does not establish better results: check the inputs, constraints, supported controls and measured performance.
Three different jobs
| Approach | What it does | What to assess |
|---|---|---|
| Energy monitoring | Shows consumption, generation and battery information. | Visibility is useful, but does it issue control commands? |
| A fixed charging schedule | Uses times and settings chosen in advance. | Can the schedule respond when prices, weather or demand change? |
| Forecast-based optimisation | Plans battery actions using expected conditions and constraints. | Are the forecasts relevant, are decisions executed reliably, and is the baseline fair? |
How to think about the economics
For a simplified battery-shifting example, assume electricity costs 10p per kWh when charging, 30p when needed later, and round-trip efficiency is 90%. Delivering 1 kWh later needs about 1.11 kWh of charging energy, costing about 11.1p. The gross difference is about 18.9p before battery wear and service fees.
These are invented example inputs, not current tariff rates or a SolarMind result. Real planning also considers solar that might otherwise be exported, available battery space and your chosen reserve. Charging at the lowest price is not automatically the best overall decision.
Examples in the market
Loop’s published guide describes consumption and solar forecasts feeding a half-hourly battery schedule. Decent Energy describes battery decisions using tariff, household and carbon information. These are provider descriptions, not independent performance tests. [1] [2]
SolarMind combines tariffs, solar forecasts and household needs in a plan designed to cover solar, battery and EV charging. It is currently onboarding suitable UK households individually for a free trial.
A useful evaluation should record
- The battery and inverter model, usable capacity and power limits.
- The exact import and export tariffs and the home’s measured energy flows.
- Your existing control method and equal starting and ending battery energy.
- Data gaps, connectivity failures, overrides and battery reserve settings.
- Net cost after fees, not only the amount of energy shifted.
No savings percentage can be guaranteed for every home. Use a representative observation period and distinguish simulated estimates from measured household outcomes.