Research · Buildings and grids · Oct 11, 2026

Power through feeders without a meter, a year nobody had read

Met40 feeders, a year held out

The test at a glance

What it tests: Accuracy against reality: power through unmetered feeders every fifteen minutes.
Data: FeederBW v1.1 (Netze BW and KIT); fitted on feeders 1-160 before 31 March 2024.
Held out: Feeders 161-200, 31 March 2024 to 31 March 2025, downloaded after registration.
Compared against: Calibrated physics; a learned model.
Bar: Lower mean normalised error than both, Wilcoxon p < 0.05 against each, and lower in every season.
Result: 0.481 against 0.681 and 0.519; ahead on 38 and 30 of 40 feeders: met.

Living Worlds predicted the power flowing every fifteen minutes through 40 real low-voltage feeders that had no meter, using the metered feeders around them, for a year it had never seen. It came out ahead of calibrated physics on 38 of 40 feeders and of a learned model on 30 of 40, with 7% lower error overall than the learned model and 29% lower than physics, and ahead of both in every season. Registered before the data was read; scored once. Data: FeederBW, Netze BW and KIT, cited with thanks.

Protocol, result and scorer

Protocol Result

Voltage at the substation as a second measure, and water networks next.

Power through feeders without a meter, a year nobody had read

More in this area.