Field notes

RDI earns executive attention when it protects money at risk: claims, disputes, delay records, and payment evidence.

Smart features become valuable when they complete a workflow that helps someone decide, act, and close the loop.

RDI starts when captured site reality becomes evidence, action, and measured outcome.

The status quo has a cost. It is not on a budget line because it is paid in fragments. Total it up before deciding the platform is unaffordable.

Meeting hygiene improves quickly when the agenda points to records rather than recollections. The minutes shrink and the actions close.

A closeout archive earns its keep in the years after handover. Keep it longer than the contract requires and the project keeps paying back.

Solar, wind, and battery storage projects share a few features that make RDI useful: distributed sites, weather sensitivity, and stringent commissioning regimes.

Residential adoption favours stakeholder reporting, progress evidence, and a tidy handover. The economics are driven by units delivered.

Commercial buildings are the bread and butter of RDI deployment: predictable shape, repeated workflows, recurring evidence demands.

Linear projects have distributed risk and distributed evidence requirements. The platform shape changes accordingly.

Data centres, hyperscale fit-out, and similar projects need command-grade evidence as a default condition. The cost of doing without it is asymmetric.

The digital lead carries the stack. They want a coherent set of layers, not another silo, and they will spot the difference within a week.

An owner representative cares about trust and reporting. RDI either makes the contractor easier to trust or invisible to the owner.

The HSE lead measures the platform by closeout rates and pattern visibility. The conversations that matter are with the regulator, the insurer, and the workforce.

The CD reads the platform as a record of arguments won and lost. The numbers that matter are payment cycles and dispute outcomes.

A project director does not need more data. They need fewer surprises and quicker triage. RDI either helps with that or it does not.

Different buyers want different things from the same platform. Reading the room is half the conversation.

A BIM model that does not get compared to reality is a record of intent. Reality data turns it into a working tool.

A payment application with evidence attached pays faster than one without. The cycle time saving compounds across the project.

Controlled access can be evidenced without a guard hut at every gate. The records do most of the work, with the right routing.

Site theft is reconstructed from records, not solved live. Know what can be recovered and what cannot before the investigation begins.

An export that holds up in legal review reads differently to one prepared for a meeting. Plan for the harder reader.

Phones, helmets, and wearables fill the gaps that fixed cameras and 360 routes leave. Plan them, do not just allow them.

360 capture is most valuable inside QA workflows, where a repeatable route turns site walks into comparable records.

Aerial surveys add a perspective the ground cannot give. Treat them as one record in a chain, not as a standalone deliverable.

Time-lapse is a useful artefact in a narrow set of conversations and almost useless in others. Use it where it earns its keep.

There is a point at which more capture stops improving decisions. Recognise it, and reinvest the budget in the layers that still need work.

RDI value comes in two shapes. Replacement workflows take cost off the project. Supporting workflows reduce risk. The two are modelled differently.

Public numbers teach a market what to expect. Project numbers describe what one project did. Confusing the two damages both.

Public ROI numbers without a band are advertising. Public ROI numbers with a band can survive the room.

RDI sits next to three large adjacent categories. Holding the boundary between them keeps each one honest.

When the same exception keeps surfacing across projects, it has stopped being a project incident and started being programme work.

A dashboard summarises. A command view triages. The two artefacts look similar on a screen and serve different purposes.

A regulator visit, an insurer query, or a coroner's inquest does not invent the evidence. The project either has it or does not.

Treat audit readiness as posture, not as a sprint when the regulator calls. The cost of staying ready is small. The cost of getting ready is not.

Handover packs assembled in the last six weeks are uniformly worse than handover packs assembled continuously. Make handover a posture, not a sprint.

A non-conformance closes when the as-built record agrees with the design intent or when the design intent has been formally amended. Anything else leaves a question for handover.

Attendance is a contractual fact. When the project can show who was on the work face, when, and for how long, the conversation is shorter and fairer.

Most delivery disputes are not arguments about the delivery. They are arguments about what nobody can prove anymore. The record removes both.

Gate logs do quiet, durable work in the commercial record. Treat them as evidence and the disputes they settle are no longer surprises.

A finding logged is not a loop closed. Closeout is a deliberate step, often the one that goes missing on busy projects.

OAC meetings change shape when the evidence is on the table. Less recollection, fewer reruns of last week, more decisions.

Substantiation lives in the cross-checks. A delay claim with a time-aligned record survives the scrutiny that ends most claims early.

A weather claim without site capture is half a claim. The other half lives on disk somewhere, if it was preserved.

Talking about avoided low-frequency, high-cost events takes a different vocabulary than talking about hours saved.

Routine progress and safety reporting is a recurring operational tax that accountants would rather not see itemised. Itemise it anyway.

Find a clip, lose an hour. Multiply by every meeting, every claim, every audit. The unindexed archive is the single largest hidden cost on most projects.

Every uncovered area on a project is a choice, even when nobody made it consciously. Treat it that way.

Plan capture by the workflows it must support. Camera count is an output of that work, not the input.

Borrow the legal concept, adapt it to site evidence, and the credibility of the record stops being a per-meeting argument.

When the captured record is not time-aligned and cross-checkable, the layers above it stop working in predictable ways.

A workflow is a closed loop, not a feature list. Two examples make the shape concrete.

Two projects, two procurement briefs, two different outcomes. The difference was not the hardware on the wall.

Capture, ground truth, interpretation, action, and command — what each layer does and what fails when one is skipped.