Skip to content
Open models
Reference modelStartup and Stabilization Delay

Startup & Stabilization Loss Model

Measure ramp losses after starts and grade changes

Paper · Glass · CementMetals

Reference implementation. Industrial use requires validation on plant-specific data. Metrics are from open evaluation splits, not customer plants.

Intended use

Measuring how long assets take to reach rate after starts, restarts, and grade changes, versus the demonstrated best ramp.

Inputs & outputs

Inputs

Start events, historian ramp curves, grade-change records

Output

Ramp loss versus demonstrated best, per start type

Training data & evaluation

Training-data classification

Synthetic ramp curves across start and grade-change types.

Evaluation approach

Time-to-rate MAE against best ramp on a held-out split.

MAE 6 min· open split

Industrial relevance

Applies across Paper · Glass · Cement, Metals. The value is highest where the affected asset is a constraint and additional output can be monetized.

Limitations

  • Requires historian resolution fine enough to resolve the ramp.
  • Best-ramp reference should reflect current equipment condition.