Introducing Lumerian Labs

01Why we started

Most of the evidence for the next decision on an energy asset already exists. It sits in a thermal flight, a title commitment, a grid study, a construction schedule or a work order. Each of these lives in a different tool, is read by a different team, and reaches the person who decides long after it was captured.

The IEA expects data-centre electricity demand to roughly double between 2025 and 2030 ^[1]. Of the US generation and storage capacity that requested a grid connection between 2000 and 2020, 13% had reached operation by the end of 2025 ^[2]. Across 373 GW of US solar, average power loss in 2025 was 5.08% ^[3]. Each raises the cost of evidence that arrives late, is read in part or never reaches the person who decides.

02What we are building

Lumerian Core keeps a persistent identity for real things (modules, parcels, permits, milestones) and records every observation of them. It turns each observation into a recommended next action, with its evidence and uncertainty attached. Field crews, developers and project teams carry out the action, and their results return to the record as labelled history that no public model has seen. Three products run on that record across the life of an asset.

• –Prism, for development: data rooms, title, public law and GIS layers, delivered as fatal-flaw and permit matrices with every statement cited.
• –Helm, for the build: the portfolio board, with every re-baseline, date change and escalation captured with its cause.
• –Aperture, for operation: radiometric thermal inspection, with every detection assigned to a module and every finding audited by an expert.

03What we hold to

1. 01Every answer carries its source. An engineer, lender or crew should act only on a conclusion they can trace.
2. 02A person signs off. The system prepares the evidence; the decision stays with someone accountable for it.
3. 03The model goes to the data. Some operators cannot connect operational systems to the cloud at all, so the same build has to run where the data lives ^[4].
4. 04Every capability runs behind a gate. Each has an experiment with a numeric bar. Below the bar, a named fallback runs, and we say so.
5. 05We publish what we learn, including what does not work.

04Where we stand

We start with utility-scale solar. Aperture covers seven site presets in six solar parks in Hungary, with 82,368 radiometric thermal frames from seven campaigns between 2023 and 2025. Campaigns in 2026 were flown at all seven sites, and the largest layout mapped holds 116,391 modules. Prism has engagements on US solar projects in Virginia, Ohio and Texas. Helm is open to its first design partner. Our first white paper, which argues for replacing risk-matrix scores with calibrated, cited probabilities in title and survey diligence, is on our research page ^[5]. If you run, build or finance energy infrastructure, write to us at hello@lumerianlabs.com.