Battery Simulator

Think manifolds, not PDEs.

The Bayesian Virtual Lab lets you sweep operating parameters and watch posterior uncertainty bands update as the MW model reasons over a battery’s degradation manifold.

What you can try

Ask “what if” before you change the setpoint

The simulator is a demonstration tool. Your own assessment uses your log; the simulator shows how the model reasons about operating levers in general.

  • Sweep operating conditionsChange temperature, state-of-charge window, depth of discharge or C-rate and watch the forecast move.
  • See uncertainty, not just a linePosterior bands widen and narrow as the model becomes more or less sure.
  • Tell real effects from noiseAn operating change is only worth acting on if it moves the forecast by more than the model’s uncertainty.
The alternative

An alternative to the 300-year-old pipeline

Classical physics simulation is rigorous, but it was never built for fast decisions on live infrastructure, so we built a faster path alongside it.

The classical way

  1. Real world
  2. Lagrangian
  3. PDEs
  4. Solve (can take weeks of compute)

The MW way

  1. Real world + physics priors
  2. A learned manifold of valid states
  3. Query (fast inference)

Try it on your own data.

Share a BMS/CAN log. We run it through the MW Battery Engine and send back an MW Battery Assessment Report: findings, how sure we are, and what to do next. The first assessment is free.