A transformer you cannot replace for three years is not a maintenance problem. It is a capacity problem, and it costs you every week it is down. We forecast which asset goes, roughly when, and what is driving it — early enough to order the part, schedule the outage, and keep the megawatts earning.
Same failure. Same part. Same price. The only thing that changed is whether the replacement was already on the dock.
Lead times, June 2026: generator step-up transformers 100–150+ weeks, substation transformers 85–110 weeks, medium-voltage switchgear 60–104 weeks — Terrapin Construction Group. PwC analysis reported by pv magazine USA in May 2026 put high-capacity transformer lead times at four years.
Asset runway is how long until a piece of equipment crosses an action threshold, together with what is driving it there and what waiting costs.
A duration, a cause and a cost, delivered together.
What is asset runway? →Every forecast answers all three, because each one pays for itself in a different way.
Which asset is actually going, out of the hundreds that all look fine on a dashboard. You stop servicing and replacing equipment with years of life left in it, and you stop spreading the maintenance budget evenly across a fleet that is not degrading evenly.
Long enough ahead to get in the procurement queue. With replacement lead times running two to four years on transformers and switchgear, this is the difference between a planned swap and megawatts sitting idle for the length of the queue.
What is driving the degradation — heat, cycling, load, a specific stress. That tells you whether you can fix it and keep the asset, whether changing how you run it buys you years, and whether the bill belongs to you or to the manufacturer.
Most systems answer none of these. Condition monitoring tells you an asset looks abnormal right now. Anomaly detection tells you a number moved. Neither tells you what happens next, or when, or what to do about it.
You buy the answer, not the architecture. Which asset, how long, what is driving it, and what it costs to wait.
Remaining-useful-life per asset. Forecasts when an asset crosses end-of-life, attributes why — and abstains when the data cannot support a call.
How one asset's degradation reshapes its neighbours' limits. System risk, not component risk — the whole facility as one connected system.
The commercial decision layer. Turns remaining life into a cost-optimal action — act now or wait, here is the window, and here is the justification record.
Degradation-aware market dispatch. Capture grid revenue without spending more asset life than the revenue is worth.
You do not have to care which is which. You care that the answer arrives as a date and a dollar figure.
Sits on top of the prediction. Turns "how much life is left" into a cost-optimal action: intervene now, or wait — and here is the window. It weighs the price of planned maintenance against the far higher price of an unplanned failure, and writes the justification record for every call it makes.
Reads the sensor signal from power-electronics assets and predicts remaining useful life — not a black box. Where physics earns its place it is built into the model, so the system cannot predict the impossible. Outputs a range with confidence, not a single guess.
You are being asked to spend money on a forecast. Here is what stands behind ours.
Any figure we put in front of you comes from a specific study on a specific date, and we will show you the study. Nothing is rounded up on the way to the slide.
On stale or incomplete data the system returns no forecast and tells you why, rather than a number that looks confident and is not. You never act on a guess without knowing it was one.
Our evidence page includes the evaluations where a simpler model beat ours. If that is the kind of thing you want to read before signing something, it is all there.
The first step runs on data you already have. No hardware, no install, no procurement cycle.
We run a read-only assessment on the history your equipment already produces and show you what the forecast would have said. If it would not have caught anything, you have lost nothing but a conversation.
When it does go live, the box sits at the asset and sends out over a private carrier connection. Your network never touches ours, which removes the review that stalls most industrial pilots.
The whole platform runs on a single edge device at the asset, so there is no data centre build-out to buy before you get an answer, and no raw data leaving the site.
One engine underneath, pointed at the asset that would actually hurt you.
The prediction only works if trustworthy data reaches it. Our partners get that data off brownfield and secured sites without rebuilding your network.
The prediction engine — the right technique per asset class, turned into dated, defensible forecasts with the misses published.
EMI-hardened WirelessHART sensing that captures vibration, thermal and pressure signals at the resolution the models require.
Cellular backhaul over a private carrier APN — traffic never touches the public internet, and your network never touches ours.
A read-only assessment proves the forecast on your own data before any hardware conversation. You see the record; you decide what it is worth.
