Key takeaways
- A claim turns on dates and sequence. Imagery is good at both, which is a different strength from the one progress photography is usually bought for.
- Imagery proves the state of a site at an instant. Everything between two captures is argued from inference, so the interval between flights is the width of the gap the other side gets to fill.
- Progress reporting is comfortable with monthly. A claim needs a cadence shorter than the disputed interval, which in practice means weekly through the contested phase plus a capture on either side of any event.
- The capture nobody has is the one before mobilization. A dated pre-work record of the site as received is the cheapest piece of evidence on the whole project and it stops being available the day the first machine arrives.
- Imagery evidences condition, extent, occupancy and access. It does not evidence cause, direction, contract interpretation, productivity or anything below grade.
- Unedited originals with intact metadata plus a short methodology record are what make a capture usable later. Annotated exhibits are built from them and kept separate, never instead of them.
- Retention and ownership have to be settled before the dispute, because a claim surfaces long after the flight and frequently after the relationship with whoever flew it has ended.
Most aerial progress capture gets bought for communication. The owner wants to see the site, the project manager wants a slide for the monthly report, and the return on progress photography is usually argued on those terms. Then something goes sideways, a change order is disputed or a schedule slips, and somebody asks whether the drone imagery can be used.
It frequently can, and it is frequently less useful than it should have been, for one reason: it was captured on a cadence designed for reporting rather than for proving anything. This is what aerial capture actually settles in a claim, what it does not, and what changes about how you buy it if the claim is the reason you are buying.
Entitlement and quantum are different questions
Any claim splits into two parts, and imagery is far stronger on one than the other.
- Entitlement. Did the condition or the event occur, and when. Was the area available to be worked. Was access blocked. Was the ground in the state the contract assumed. This is where dated imagery is genuinely good, because it is a contemporaneous record made by something with no stake in the argument.
- Quantum. How much time or money. Imagery contributes measurements, which is real but partial. A quantity from a capture is only an input to a cost; the rates, the productivity assumption and the critical path analysis all sit outside it.
Where imagery gets oversold is the step from a fact about the site to a conclusion about responsibility. A capture can show that an area held standing water on the fourteenth. It cannot show why, who was supposed to have drained it, or whether the contract put that risk on the owner. That is a document and testimony question with imagery as corroboration.
What a repeating capture actually settles
In ascending order of how hard it is to argue with:
- Site condition on a date. Surface water, snow, mud, frost, debris, stored material, vegetation, erosion. These are the facts that disappear fastest and get remembered differently by everyone.
- Extent of work complete. Which areas were graded, which were paved, where the pipe trench reached, how many units were framed. This is where sequence gets established.
- Access and occupancy. Who had the laydown area, where the crane was set, which haul route was in use, whether the gate approach was clear. Occupancy disputes between trades are almost impossible to reconstruct from the ground and are obvious from above.
- Quantities. Earthworks volumes, stockpile positions and surface areas, which is the same work as an aerial quantity for a progress payment and benefits from the same controls.
- As-built against design. Whether what is on the ground matches the issued surface, which is a surface against surface comparison and has its own set of traps.
Equipment and crew presence deserve a separate warning. A capture shows what was parked where at one moment. It does not show whether anyone was working, and it is a short step from an overhead photo of idle machines to an argument about productivity that the photo cannot support. One frame of a lunch break proves nothing about a shift.
Cadence is the whole ballgame
Imagery is a record of instants. Between two captures the record is silent, and the other side gets to propose what happened in the silence. If captures are monthly and the dispute is about a nine-day delay, the imagery brackets the delay without resolving it.
A cadence built for a claim looks different from one built for reporting:
- Weekly through the phase where the risk sits. Earthworks, underground, anything weather-exposed, anything with a sequencing dependency on another trade.
- Before mobilization. The pre-work baseline, discussed below.
- At every milestone the contract names. Substantial completion of a phase, a handover between trades, a hold point.
- Immediately after any weather event or stop-work. Within a day if the site can be flown at all. Conditions after a significant rain, hail or wind event change within hours.
- Either side of any disputed change. If a change is directed on Tuesday, the capture on Monday and the capture on Friday are worth more than any number of monthly ones.
The expensive mistake is reactive capture. A flight ordered after the dispute starts documents the site as it is now, which is the one date nobody is arguing about. There is a related discussion of frequency in the piece on how often to re-fly a site, with the difference that a claim sets the interval from the length of the disputes you expect rather than from the rate the site changes.
The pre-work baseline
Site condition claims run both directions. A contractor argues the ground was not as represented. An owner argues the damage was done during the work. Both arguments are decided by what the site looked like the day before mobilization, and on most projects nobody has that.
A single capture of the site as received, dated and retained, costs one mobilization and answers a question that otherwise reduces to competing recollections. It covers existing damage, existing drainage and standing water, existing stored material, existing access condition, the state of adjacent property and the position of visible surface features. It is the same logic as a pre-tender capture in a bid package, used defensively instead of commercially.
What makes a capture hold up
The general principles are the same ones that make any dated aerial capture defensible, and they are worth restating in a claim context because this is where they get tested:
- Unedited originals. Keep the raw frames with their metadata intact. An orthomosaic is a processed product and a processed product invites questions about the processing. Both are useful; only one is primary.
- Metadata that survives. Capture date and time, position, and the processing record. Re-exporting imagery through a presentation tool strips this, which is how a good capture becomes a weak exhibit.
- A methodology record. Flight parameters, equipment, ground control, software and version, and accuracy testing against independent checkpoints for anything measured. Short, and written at the time.
- Exhibits kept separate. Annotation, markup, highlighted areas and measurement callouts are an argument built on the evidence. Produce them as a derived document that cites the original file, never as a replacement for it.
- A chain nobody has to take on trust. Who captured it, where it has been stored, and who could have altered it. A contractor flying its own site is still useful; a third party capture with its own records is harder to attack.
Ownership and retention, settled early
A claim surfaces late. Limitation periods run for years, and by the time the file is live the project team has dispersed, the drone contractor may have changed and the original data may be somebody else's property. Three things belong in the agreement at the start:
- Who owns the raw data and who holds a copy, which is the core of the data ownership and retention question. Hold your own copy of the originals rather than relying on a vendor's archive.
- Retention period long enough to cover the period in which a claim can be brought, which is longer than most retention defaults and longer than most projects think about.
- Who may use it. If the capture was commissioned by a general contractor and an owner wants to rely on it, or the other way around, the licence wording decides whether that is possible. On a consultant commissioned capture the chain can be longer than anyone assumes.
The limits, stated plainly
Aerial capture does not see below grade, inside a structure or under water. It does not establish who directed a change, what a clause means, whether an instruction was given, what a crew was doing, or what the critical path was. It will not substitute for a scheduling analysis and it is not a general purpose site record. What it does is remove whole categories of factual argument from the table, so that the dispute runs on contract and causation rather than on whether the yard was under water on the fourteenth.
Anything touching the legal strategy, the notice requirements under the contract or the form the evidence has to take belongs with the party's counsel. Our part is the capture, the measurement and the record that makes both readable years later.
