Product
Yunify: physics-driven operating intelligence for energy plants
Yunify™ runs alongside the control system a plant already has and turns operating data into a prescribed action. It does not replace a DCS, a PLC or a SCADA layer, and it does not require them to be modified.
The components
Yunify is deployed as a set of parts. Each is useful on its own, and the value compounds when they run together.
Yunify CVM
Cell-level voltage measurement for electrolyser stacks, across hundreds of cells in series. Specific to green hydrogen, where few plants currently instrument at that level; the rest of the components are asset-agnostic.
Yunify Edge
An on-premise server that acquires data from any control system over OPC UA, Modbus, MQTT or PROFINET, with no modification to the installed PLC, DCS or SCADA.
Yunify Engine
First-principles models of electrochemistry, thermal behaviour and flow, running alongside AI models performing anomaly detection and residual learning.
Yunify Gateway
A protocol gateway that speaks OPC UA, OPC DA, Modbus TCP, MQTT with Sparkplug and PROFINET, and normalises them into one tag model.
Yunify Historian
A time-series store on the site operations network that keeps the raw record at full resolution alongside the plant state that explains it.
Why physics sits in front of the data
Purely data-driven models fail outside the history they were trained on, and renewable-coupled assets sit outside it constantly. An electrolyser following a solar profile spends most of its life at partial load and in start-stop cycles, which is precisely where efficiency loss, purity excursions and accelerated degradation originate.
In Yunify the AI detects an anomaly and the physics decides whether it is physically possible. That filtering is what removes the false alarms that cause operators to stop trusting statistical systems. It also lets the Engine estimate variables no sensor reports, including degradation, gas crossover, state of health and remaining useful life.
Because every insight traces back to a mechanism, it can be explained to an operator, and to an auditor.
The output is an instruction
Most monitoring tools report that efficiency is down three per cent. Yunify is built to report the problem, the validated cause and the prescribed action together. That means which component is drifting, why, how far it has moved from its baseline, and what to do about it. For example, on a green hydrogen plant: which cell is drifting, that its electrolyte is contaminated, how far its voltage has moved from baseline, and that the answer is a regeneration cycle and a current trim.
Three things happen in real time. Process parameters are optimised against current conditions. Failures are predicted before they force a shutdown. Production is scheduled around measured degradation rather than a nameplate curve.
Asset classes
One architecture covers the renewable value chain, from 1 MW to 1 GW. Every one of these assets is managed in aggregate while it degrades at cell, module or component level. A battery management system reports state of charge but not cell-level state of health, exactly as a DCS reports stack voltage but not cell voltage.
Green hydrogen
Alkaline, PEM and AEM electrolysers, and the balance of plant around them.
Battery storage
An advanced EMS layer: cell-level state of health, augmentation planning and cycle-depth accounting feeding when to charge and when to despatch, so the schedule answers to measured health and its financial consequence rather than to nameplate.
Solar and wind
Separating soiling from degradation, clipping and curtailment.
Conventional gas power
Heat rate degradation, compressor fouling and equivalent operating hours.
Downstream
Compression, storage, ammonia and offtake under variable supply.
From a drifting component to a number the business recognises
A finding is only worth what somebody does with it, and the four people who have to act on it do not need the same sentence. Yunify is built to carry one mechanism through all four without losing what caused it.
Operator
The action, on shift: which unit, what to change, and how long there is to do it. Nothing at the panel should require the analysis to be interpreted first.
Engineering
The mechanism and the evidence under it: which signals moved, against which baseline, and what physically explains them. This is the layer that confirms or rejects the diagnosis, so it has to be inspectable.
Planning
Lead time. A degradation rate with a date attached is what turns a repair into a scheduled outage rather than an unplanned one, and what lets spares and labour be ordered instead of expedited.
Management
The same finding in the units a budget is written in: energy per unit produced, availability, and the production deferred if the condition is left to run another quarter.
What it is designed to deliver
The operating outcomes a deployment is scoped and engineered to reach, against the plant's own baseline.
2 to 5% lower specific energy
Target reduction in energy per unit produced, such as kWh per kg of hydrogen.
Up to 5% higher yield
Target improvement under variable load.
20 to 30% fewer outages
Target reduction in unplanned stoppages.
2 to 6 weeks early warning
Target lead time on failures in the primary asset and its balance of plant, such as an electrolyser stack.
Frequently asked questions
Does Yunify replace our control system?
No. Yunify runs alongside the existing PLC, DCS or SCADA and reads from it. Control logic is not altered.
Where does Yunify run?
On premises, in the cloud, or as a hybrid of the two. The deployment mode is a choice, not a constraint of the product. On-premises is the common case: a single industrial host per plant, with no dependency on an outbound connection, which for assets with public or sovereign participation is usually a precondition rather than a preference. Where a connection is available and wanted, a cloud or hybrid deployment lets a fleet be seen together while the plant-critical path stays local.
Does it work on a brownfield, multi-OEM plant?
Yes, and on greenfield equally. Yunify is built for heterogeneous, mixed-OEM environments and reads the instruments a plant already has, so the addressable base is every plant already operating rather than new build alone.
How is Yunify priced?
An annual subscription scaled to the capacity being monitored, plus hardware and engineering services depending on scope. Figures are quoted per site, because the mix of assets and the deployment scope vary. Get in touch for a quote.