Predict the response
to demand.
One recorded electrical pulse. Four heavier loads.
Change the load, then reveal what the bench measured.
Diagnostic intelligence for equipment makers.
Recorded case demo. Saved outputs from the real model.
Ask about a heavier load.
Five-second discharge pulse on the same calibrated cell family.
Loading the recorded input…
Recovery after load removal
Voltage rise / mVEffective resistance
— mΩ
Voltage drop under this load
— mV
Calculated from requested current × estimated resistance.
More current demands more of the cell. Resistance helps explain the voltage drop that a charge indicator cannot.
The heavier-load measurements stay out of the model’s inputs.
All four load conditions use the same adapted core.
Measured on unseen batteries.
81 physical cells across three commercial families. Results below use the independent test’s three primary load levels, with equal weighting by family and cell.
| Model | Recovery / mV | Resistance / mΩ |
|---|
Loading benchmark evidence…
Study context and limitationsThis evaluation uses public laboratory cells and a calibrated recording protocol. The equipment scene shows the intended application. These estimates do not establish pack readiness, safe operating limits or military field performance.