For banks & institutional lenders
How can satellite evidence support a construction loan drawdown review?
Dated satellite observations can add an independent view of visible site change to a construction drawdown review. A lender still needs the relevant milestone definition, project boundary, documentary evidence and technical assessment. CRI is being developed to keep these inputs, their limits and the next review question together in a versioned Evidence Pack.
CRI is currently in controlled-case Proof of Concept validation. These are public explanations of the intended workflow; institution-specific use requires agreement and human review.
Start with the declared scope
In a model construction-finance review, a borrower presents a progress declaration as part of a drawdown request. The reviewer needs to understand whether the supporting material covers the same works, phase and period. A visible change on the site may help frame the follow-up, but the contractual milestone can include work that no overhead image can reveal.
The intended CRI workflow separates the declaration, the observed change and the reviewer’s response. The question should be specific enough to test, such as whether dated observations support proceeding to documentary review of a declared construction stage.
What the reviewer needs
- Defined area: an institution-confirmed asset boundary and the relevant phase or works. A working image retrieval envelope alone is insufficient.
- Relevant dates: the declaration date, the comparison period, actual acquisition dates and the review date.
- Authorised records: the applicable milestone definition, baseline, progress declaration and supporting technical records with their versions.
- Review responsibility: who can provide a scoped technical response and who accepts it under the institution’s existing process.
Source selection and review cadence depend on the agreed question and usable evidence. A fixed reporting interval does not guarantee a suitable observation for each milestone.
Fischamend: a bounded public illustration
The public Fischamend Evidence Pack compares optical observations from 20 August 2023 and 19 August 2025. Its banking question is synthetic. It does not represent a real drawdown request, bank decision or customer pilot.
| Part of the record | What it says |
|---|---|
| Declared | A hypothetical staged review question expects large roofed footprints to be visible by the later date. This is not an actual contractual milestone. |
| Observed | Two large roofed structures became visible in the selected observation area, replacing predominantly bare or construction-affected surface. |
| Uncertain | The exact project boundary, engineering completion and payment eligibility remain unverified. |
| Next review step | Request the dated documentary and engineering records needed to examine quantities, acceptance criteria, environmental conditions and institution-specific payment conditions. |
WATCH in this record means evidence sufficiency needs documentary review. It is not a negative project rating and does not indicate delay, breach or default.
Source: Fischamend Evidence Pack, DEMO-EU-AT-FIS-01, v0.1 public-safe release. The full record retains the source identifiers, observation envelope, reasons and qualifications. No affiliation with or validation by the project owner or financing institutions is implied.
Where the image-based conclusion stops
Visible roofed footprints can support a narrowly defined observation. They do not establish building completion, workmanship, fit-out, occupancy, environmental compliance, contractual acceptance or eligibility for payment. Documentary and engineering evidence remain necessary before any institution-specific conclusion.
Resolution, cloud and shadow can also limit what is observable. The guide to satellite evidence limits explains how these constraints affect a review question. A scoped acceptance of one technical response does not authorise a drawdown.
Discuss a specific review question
Bring the asset context, the available source types and the question a responsible reviewer needs to answer. An Evidence Readiness Check can examine whether the evidence and workflow are suitable for a bounded evaluation.
3BrainAI Nexus · Published