What matters
The problem comes before the model.
- Detect anomalies and drift without confusing a model with process truth.
- Estimate quality or properties between lab assays while declaring when the model should abstain.
- Compare energy or dispatch alternatives against an explicit baseline.
Lab Intelligence approach
Sources, assumptions, checks, versions, metrics and limitations remain visible. A more complex technique is used only when it adds value over an understandable baseline.
How the method works →Related workflows
Examples already documented.
These pages describe implementations or experiments from the laboratory; they do not imply that the same result transfers automatically to another context.
Process Sentinel
Detects anomalies and faults on a generic dynamic chemical process using a frozen 28-day synthetic data (simulated data) benchmark, isolated truth and a drift-aware PCA + process-residual ensemble.
Open project →WF-15Virtual Analyzer
Experimental soft sensor on synthetic data (simulated data): estimates product quality between laboratory assays with delay, noise, point-in-time controls, OOD detection and measurable abstention.
Open project →WF-16Energy & Steam Optimizer
Optimises a modelled HP/MP/LP steam network with utility demand based on synthetic data (simulated data) and real/configured gas, power and carbon prices against a frozen reference dispatch policy.
Open project →WF-11Hydrogen Route Observatory
Compares grey SMR, SMR+CCS and electrolysis using public/configured energy prices, chemical balances, emissions boundaries and break-even thresholds.
Open project →WF-12Ammonia & Fertilizer Chain
Propagates gas cost through the NH₃→urea chain while separating the physical calculation, nowcast and falsifiable M+1 forecast.
Open project →Want to discuss a similar case?
Describe the problem and expected result in non-confidential terms. The first step is to identify the data, criteria and limitations that would make the case verifiable.