A model of the line that knows why a reading changed.
Physics-based digital twins, maintenance forecasts with the mechanism shown, vision on every unit, and one identity per asset across SCADA, historian, CMMS and ERP.
- 01Physics behind the dashboard
- The instrumentation is usually there. We add a model that links the readings to the thermodynamics and mechanics of the process, so each change has a physical cause.
- 02Forecasts with reasons
- Every maintenance recommendation includes the wear mechanism and the confidence behind it, so a planner can act on it and an audit can trace it.
- 03Every unit, every shift
- Camera inspection of every casting, weld or pack, and of the safety zones a person cannot watch all day.
01Pain points
What we help solve
- 01
Predictive maintenance that stopped predicting
Models fitted to a snapshot raise false alarms or miss failures as feedstock, load and ambient conditions move.
- 02
Four systems, no single picture
SCADA, the historian, the CMMS and quality each with their own data model, and no single record that links them.
- 03
Inspection by sampling
Quality checks and safety monitoring by people who cannot be at every station, so a sample stands in for the whole.
- 04
Dashboards without a model behind them
Sensors installed and a platform bought, but nothing that connects the dashboard to the physics of the process.
02Capabilities
Our work in manufacturing
A model of the line, forecasts with reasons, vision where people cannot be, and one name per asset.
- Physics-based twins of the line
From a single pump to the whole line, calibrated on your history.
- Hierarchical models from equipment to production line
- Calibrated on operating history, governed by the physics
- Diagnosis of an anomaly with a recommended action and its reasoning
- Accuracy maintained through process drift
- Maintenance forecasts with the mechanism shown
Every recommendation names the wear mechanism behind it.
- Equipment-specific degradation models for rotating and static equipment
- Remaining useful life tied to the mechanism
- Operator-facing explanations for every recommendation
- Condition-based scheduling
- Vision on the line
Every unit checked.
- Defect detection and quality verification
- PPE and exclusion-zone monitoring
- Visual anomaly detection on process equipment
- One system across many sites
- Plant data reconciliation
One identity per asset across SCADA, historian, CMMS and ERP.
- Entity resolution across the plant's systems
- Provenance on every reconciled value
- Ambiguity routed to a person
03Questions
Questions we take on.
01Why has the reformer's efficiency dropped two per cent since March, and is it fouling or instrumentation?
02Which pumps should be opened at the next turnaround, and which can wait?
03Can we check every casting on the line?
04Solutions
Solutions we draw on
All solutions01Industrial AI
Industrial AI and automation
First-principles digital twins, forecasts with the reasoning attached, computer vision and autonomous systems for plants, fleets and cities, with the cause and the confidence shown next to every number.
02Simulation
Engineering simulation
Custom and open-source solvers for design, certification and optimisation. We cover fluids, structures, heat and electromagnetics, coupled where the physics is coupled, on infrastructure you control.
03Research
Scientific research and development
Model formulation, discretisation, convergence analysis, physics-informed learning and foundation models for simulation, built to run at industrial size.
04Software and HPC
Software engineering and HPC
Parallelisation, GPU kernels, simulation pipelines, testing and deployment for numerical software that has to run in production.
Contact