0
buses modelled live
Physics-grounded · Bayesian
See the true state of your grid — and how far to trust it.
GraphSolve Power turns noisy grid data into a clear, live picture of your network — and puts an honest confidence level on every number. Spot failing sensors, catch problems early, forecast demand, and relieve congestion, all in one place.
your grid, livereading sensors…
0s
to read the whole grid
±σ
confidence on every value
0%
of the network, one model
Why it matters
Most AI gives you one confident answer — and no warning when it’s wrong.
On a power network, “confident and wrong” is the failure that trips a system. GraphSolve Power is built the other way round: grounded in the real physics of your grid, and Bayesian to the core — so every answer arrives with a calibrated confidence you can act on.
Ordinary AI — one line, no warning
GraphSolve Power — an answer that knows its limits
Confidence on every number
Voltages, forecasts, sensor health — each comes with a clear confidence level. A model that tells you when it isn’t sure is a model you can actually run operations on.
It knows which data to trust — and which would help most
Feed it a new reading and it updates its whole picture of the grid, showing exactly where confidence improved. It even tells you which extra measurement would sharpen things most — so you invest where it counts.
Grounded in your grid, not a generic guess
The real physics of your network is built into the model, so its answers stay sensible even where data is thin — and only get sharper as more data arrives.
Measured, not claimed
Confidence you can check.
We put it to the test against the standard estimator used across the industry — on the IEEE-118 benchmark network, the same data into both. Same answer. Only ours tells you how far to trust it.
Calibrated on every value. Both methods return the same estimate — but GraphSolve adds a confidence interval that brackets the truth 95% of the time, and whose width tells you which points to trust. The standard method hands you a bare number.
Built for real time, at scale. A calibrated confidence on every value, engineered to run faster than the standard estimator — and to scale from a single substation to a national grid.
The difference
Planning tools size the grid you might build.
We tell you the truth about the grid you’re running.
One live, trustworthy model of your whole network — the layer that sits between your sensors and every decision you make on them.
What it does
One model. The whole operating picture.
Explore each of these yourself in the live web demo — running on a real 585-bus national grid.
Live state
The true state, with confidence
Turns thousands of noisy readings into one clear picture of the grid — every value with a confidence band, and bad data flagged automatically.
585 buses · in seconds
Sensor health
Which meters to trust
Spots sensors that have quietly drifted out of calibration — before their bad readings mislead your team.
whole grid · instantly
Anomalies
Catch what shouldn’t be there
Flags unusual patterns in live measurements and in event and alarm streams, ranked by how unusual they are.
signals + events
Forecasting
Demand ahead, with confidence
Forecasts load with honest confidence bands that widen the further out you look — so you can plan around the risk, not just the number.
24 h · calibrated
Congestion relief
Relieve overloads, live
Finds the overloaded line under peak demand and recommends the cheapest way to bring it back within safe limits.
115% → 100%
Coming soon
Always-on monitoring
The live state, its confidence, and drift and anomaly alerts — updating continuously from your real-time feeds.
on the roadmap
Works with your data
Plug in your network.
GraphSolve Power speaks the open standards the industry already uses, including the PyPSA modelling ecosystem. Point it at your network — or a public one — and the full picture comes to life. Our live demo runs on a real national transmission grid out of the box.
PyPSA SciGRID PyPSA-Eur MATPOWER CIM your data
Our view on AI
Give AI better tools —
don’t dress its guesses up as physics.
AI is a superb interface. It makes powerful systems approachable and puts an expert at every operator’s side — and we lean into that: ask our assistant about your grid in plain language. But an interface is not a source of truth.
Where rigour — and sometimes lives — are on the line, the right way to bring AI into the control room is to give it real, physics-grounded tools to reach for, not to wrap a model’s output in more machine learning and call it physics. A neural network trained to emulate a grid is still a guess, and on a live network a confident guess is the dangerous kind. Our AI reaches for a calibrated model of your actual network — and reports what it truly says, uncertainty and all. The AI is the copilot; the physics is the source of truth.
How we compare
A different job entirely.
Planning & optimisation tools
GraphSolve Power
What it answers
What grid to build next
What your grid is doing right now
Certainty
One answer, take it or leave it
A confidence level on every value
Grounding
A generic statistical fit
Built on your grid’s real physics
Data quality
Assumes the inputs are clean
Catches bad data & drifting sensors
Timeframe
Long-range planning
Live & operational
Where we sit in the wider landscape
Everyone models the grid. Almost no one gives you a live, physics-grounded state with a calibrated confidence on every value — the top-right corner.
| Capability | GraphSolve Power | EMS / ADMS incumbents | Open-source SE | Physics-AI platforms | Forecast vendors |
|---|---|---|---|---|---|
| Real-time, physics-based state | ✓ | ✓ | ✓ | — | — |
| Calibrated confidence on every value | ✓ | — | ◐ | ◐ | — |
| Bad-data detection | ✓ | ✓ | ✓ | — | — |
| Sensor-drift detection | ✓ | — | — | — | — |
| Anomaly detection (signals & events) | ✓ | ◐ | — | — | — |
| Probabilistic forecasting | ✓ | ◐ | — | — | ✓ |
| Congestion relief | ✓ | ✓ | ◐ | ◐ | — |
| Open-standard / PyPSA interoperability | ✓ | — | ✓ | — | ◐ |
✓ full · ◐ partial · — none Categories, not specific vendors.
Who’s behind it
Proven physics-AI, now for the grid.
GraphSolve Power is built on the Data Insights AI engine — the same engine behind GraphSolve Energy, already trusted on some of the most complex physical systems in industry.
PM
Dr Peter Mann
Founder. A renowned physicist, mathematician and graph-theory pioneer — computer scientist and lecturer, and author of Lagrangian & Hamiltonian Dynamics (Oxford University Press, 2018). The research mind behind the engine.
AT
Dr Alan Tominey
Founder. Scientific software developer, numerical-modelling specialist and optimisation expert — the engineering depth that turns the research into a product operators can run.
Trusted on industrial systems in the North Sea · GCC · LATAM
Get started
See the true state of your grid.
A live, web-based demo on a real national grid — no install. Walk through it yourself, or let us show you.