Rendered PEM electrolyser stack. Illustrative.

HYDRA OS platform

One engine. Four ways to use it.

HYDRA OS is the independent performance intelligence platform for water electrolysis. Assess, Verify, Monitor and Review all run on the same electrochemical engine.

Talk to our team See the architecture

Stage 1 · Data quality and normalisation

Checked before it is trusted.

Every dataset is checked for gaps, drift and stuck sensors, steady-state windows are found, and values are corrected to reference conditions before any comparison is made.

Stage 2 · Electrochemical engine

Physics that can be inspected.

A deterministic engine covering reaction kinetics, ohmic and transport losses, thermodynamic balances and degradation. Each module is an equation or a balance you can read.

Stage 3 · Uncertainty and cross-checks

Every finding carries its uncertainty.

Independent modules cross-check one another before a finding is released. Outputs become KPIs, reports, verdicts and alerts, each with its uncertainty and its sources.

Continue to the page

Where the engine reads

Skip

From data to evidence.

Six stages, the same for every product. What changes between products is the question asked, not the engine that answers it.

HYDRA OS data flow

The engine.

At the core of HYDRA OS is a deterministic electrochemical engine: a large library of computational modules, each an equation or a balance that can be inspected. They cover reaction kinetics, ohmic and transport losses, thermodynamic balances and degradation. Independent modules cross-check one another before any finding is released, and every finding carries its uncertainty. Language models never perform calculations: they only draft report text from results the engine has already computed.

The product family.

HYDRA OS product family and buyer roles

HYDRA Assess

Performance and uncertainty from test or operating data.
HYDRA Verify

Guarantee verification with a clear verdict.
HYDRA Monitor

Continuous, read-only, on-premise.
HYDRA Review

The reviewer toolkit.

What HYDRA OS does.

The capabilities behind the four products, on one engine.

Capability	Products
Electrochemical engine calibrated against NREL benchmark data	Assess · Verify · Monitor · Review
Data-quality checks	Assess · Verify · Monitor
Polarisation curve fitting and performance baseline (kinetic, ohmic, transport losses)	Assess · Monitor
Normalisation of operating data to reference conditions	Assess · Monitor
Correction of acceptance tests to guarantee conditions	Verify · Review
Guarantee verdict (PASS / FAIL / INCONCLUSIVE)	Verify
Degradation tracking, separating reversible from irreversible losses	Assess · Monitor
An uncertainty range next to every figure	Assess · Verify · Monitor
Projection under the intended operating profile	Assess
Short-stack to full-stack scale-up	Assess
Representativeness score (scale, stack generation, operating envelope)	Assess
KPIs complementary to DNV-RP-J302 performance tests	Assess · Verify · Review
Remaining useful life estimate and stack replacement planning	Assess · Monitor
Continuous on-premise, read-only monitoring	Monitor
Uncertainty that narrows as operating data accrues	Monitor
Early fault warnings, ranked by confidence	Monitor
Specification normalisation across OEM quotes	Review
Report text drafting (language model, from computed results only)	Assess · Verify · Review

Data requirements.

What we need

Timestamped stack voltage and current (or current density), stack temperature and pressure. Cell voltages, gas purity (H₂ in O₂), electrolyte or water conductivity, and factory or site acceptance test curves improve the result but are optional. A log of shutdowns and maintenance events helps separate reversible from irreversible losses.

Formats

CSV or Excel exports from your historian or test bench. Other formats on request.

How it reaches us

Assessments: as files, with no connection to your plant. Monitoring: read-only from your historian or control network, integrated during commissioning.

Deployment.

From data files

Assessments run on data files you export, with no connection to your plant. Nothing writes to your control system.
On-premise monitoring

HYDRA Monitor runs inside your network, read-only, with no inbound connections. Air-gapped operation is confirmed for your site during scoping.

How we handle the data you send us: Trust Center

Questions.

Our OEM already monitors the stacks.

Keep it. HYDRA OS gives you an independent reading of the same data, which is what lenders, insurers and warranty discussions call for.

Is it safe to connect this to our plant?

HYDRA OS is read-only and advisory. It never writes to control or safety systems. Assessments need no connection at all, only an export.

Our data is confidential.

It stays on your site, or with us under NDA during an assessment, and it is deleted afterwards with written confirmation.

How do we know the numbers are right?

Every KPI follows a published definition, every figure carries its uncertainty, and every report records the data quality and the engine version used.

This asset runs for 20 years or more.

Open export formats, a documented data model and source-code escrow on request. Your evidence stays usable whatever happens.

Start with an assessment.

Send us an export. Our team returns verified KPIs, degradation rates and a clear view of where your voltage goes.

Request an assessment Talk to our team

One engine. Four ways to use it.

Checked before it is trusted.

Physics that can be inspected.

Every finding carries its uncertainty.

From data to evidence.

The engine.

The product family.

HYDRA Assess

HYDRA Verify

HYDRA Monitor

HYDRA Review

What HYDRA OS does.

Data requirements.

Deployment.

From data files

On-premise monitoring

Questions.

Start with an assessment.