The first capture of a new site costs more and takes longer than any subsequent capture of the same site. Recognising why turns the first project from a frustrating overhead into a strategic investment that locks in efficient delivery for every cycle that follows.
If you've ever quoted the second capture of a site at 40% below the first and watched a client wonder why the original was so expensive, you've met the first-capture economics. The first project of any new site carries setup work — control, calibration, flight planning, processing pipeline configuration, documentation — that the second project doesn't. Once those investments are made, every subsequent capture amortises them.
This article is the buyer-facing version of why that gap exists, what gets locked in, and what to specify on the first contract so the rest of the programme runs at steady-state cost.
Seven things get locked in by the first capture of any new site:
1. Ground control regime. Where the permanent or hot-tied control marks sit, what AHD heights they're tied to, what distribution they have across the site. Established once; referenced every subsequent capture.
2. Flight plan. The takeoff/landing points, flight altitude, swath geometry, overlap, total airtime. Optimised for the specific site on the first capture; reused on every subsequent capture.
3. Processing pipeline configuration. Classification parameters tuned to the local vegetation and terrain types, hydro-enforcement breakline list, sensor calibration values. Tuned once; runs identically every cycle.
4. Deliverable format spec. LAS version, LAZ compression, tile structure, naming convention, manifest format, surface grid resolution. Defined once; reused without renegotiation.
5. Datum and CRS choices. GDA94 vs GDA2020, MGA zone, AHD, project local grid if applicable. Locked in for consistency across all captures.
6. Tile structure conventions. Grid alignment, tile size, naming pattern. Established to match the project's downstream tools and the operator's processing pipeline.
7. QA acceptance criteria. What residual RMSE is acceptable, what classification accuracy is required, what the sign-off threshold looks like. Documented once; applied to every cycle.
A capture that runs without locking these in produces a fine single deliverable and an awkward second project where every decision has to be re-litigated.
Three categories of work happen on the first visit and don't need to happen again:
Ground control installation. If the site doesn't have suitable permanent marks, the first capture is where they get installed and surveyed. This is licensed-surveyor work, sometimes contracted separately, sometimes bundled. Either way it's a one-time cost.
Geodetic control adjustment. Tying the installed marks to AHD/MGA via observed GNSS over hours-to-days, processed against CORS or commissioned base station. The result is a project-specific control file that every subsequent capture references.
Site reconnaissance and stakeholder coordination. First visit involves walking access, identifying staging areas, coordinating with landowners or operators, validating airspace. Second visit goes straight to capture.
For mine sites or industrial facilities, the first visit also covers safety induction, SSE coordination, PPE provisioning, permit-to-work setup — all of which are documented once and reused.
The flight plan that's loaded into the drone on day one is almost never the final flight plan for the programme.
The first capture surfaces:
The first flight delivers the data; processing the data informs the second flight plan. By the third capture, the plan is stable and produces predictable results without further iteration.
Standard processing software has parameters that need tuning to specific site conditions:
Classification parameters. CSF cloth resolution and rigidness, progressive TIN angle thresholds, vegetation height bins. Defaults work for typical conditions; specific sites (dense bushland, steep terrain, structures-dense urban) need parameter adjustments that emerge from the first project.
Manning's n / land cover classification rules. For hydraulic and vegetation work, the rules that classify cover types from LiDAR + ortho need calibration against the site's actual vegetation and surface types.
Hydro-enforcement breakline list. Bridges, culverts, weirs, pipes — every structure that affects hydraulic flow needs an explicit breakline. The first capture documents the structure inventory; subsequent captures reuse it.
Sensor calibration values. Boresight calibration applies across captures but periodic re-validation against the site gives confidence the values are still good.
Once tuned, the pipeline runs identically on subsequent captures — same classification quality, same surface model, same hydro-enforcement, same outputs without per-cycle calibration overhead.
The first QA pack is also more substantial than subsequent packs:
Subsequent QA packs reference the first as the baseline, then record the cycle-specific updates. Much shorter, faster to produce.
Realistic cost differentials, first capture vs steady-state subsequent:
| Item | First | Subsequent | | --------------------------------- | ------------- | ----------- | | Site reconnaissance + coordination| 1-2 days | None | | Control installation | $3k-15k* | None | | Geodetic adjustment | Half day | None | | Flight plan + permits | 1-2 days | Already done| | Processing pipeline configuration | 1-2 days | Re-use | | Standard capture + processing | Per-project | Per-project | | QA documentation | Full pack | Delta only |
*Control installation cost varies — sites with existing permanent control nearby pay nothing; remote sites needing new permanent marks pay the full installation + survey cost.
For typical mid-sized engineering captures, the first cycle runs 30-80% above the steady-state per-cycle rate. Programmes of monthly or quarterly captures amortise the differential across many projects.
Worth structuring contracts to reflect the differential explicitly — quoting "first capture $18k + subsequent $11k each" is more transparent than "average $13k per capture".
Five things worth specifying on the first contract to make subsequent cycles cheap:
1. Permanent ground control. Pay for the marks to be installed properly. They benefit every subsequent capture and support any future engineering work on the site.
2. Flight plan documentation. Insist that the flight plan gets formally documented (waypoints, altitudes, parameters) as a project deliverable. Without it, the second cycle reinvents the plan.
3. Processing pipeline version. Capture the specific software versions and parameter sets used. Re-processing against an updated pipeline is fine; doing it inadvertently because nobody documented the original is not.
4. Deliverable format + tile structure. Locked in the first cycle; reused identically every cycle. Variability between cycles defeats the purpose of recurring capture.
5. Independent reference data (where applicable). For stockpile work, the cleared pre-pile ground capture. For environmental work, the pre-disturbance baseline. For infrastructure work, the as-built reference. The first cycle is the only time these can be captured.
For some project types, the first capture is the most valuable single capture rather than just an expensive prelude:
Pre-disturbance baseline for environmental sites. The ground as it was before the project started. Reference for every subsequent compliance check, rehabilitation milestone, closure submission.
Pre-construction baseline for civil sites. The terrain before earthworks. Reference for cut/fill computation, design validation, claims documentation.
Pre-placement ground for stockpile reconciliation. The cleared pad before piles go down. Reference for every subsequent volume calculation that doesn't have to assume a base plane.
(See mining stockpile LiDAR — the base-plane problem for the volumetrics version of this.)
Pre-event baseline for flood-affected sites, mine operations, vegetation programmes. Whatever the project needs to compare future states against.
For all of these, the first capture isn't an overhead — it's the foundation that makes every subsequent capture meaningful. Worth investing in.
Three patterns we see when buyers don't recognise the first capture as foundational:
Treating the first capture as a one-off. Single project, single deliverable, no thought to recurrence. Then a second capture is needed 18 months later and the operator has moved on, the control is gone, the flight plan is undocumented, the processing pipeline has changed. Effective re-do at full first-capture cost.
Skipping permanent ground control to save first-capture cost. "We'll just use GNSS-only." Saves $5k on the first capture; costs the same $5k plus a half-day's lost capture time on every subsequent visit because there's no permanent reference to tie to.
Not documenting processing decisions for repeat. Operator runs the pipeline, ships the deliverable, doesn't write down which parameter values were used. Six months later when a re-process is needed, the documentation isn't there. Reproducibility is lost.
The fix in all three cases is the same: structure the first contract to deliver permanent setup, not just a one-off output.
First capture of any new site costs more than subsequent captures. Seven things get locked in (control, flight plan, pipeline, format, datum, tile structure, QA criteria); three categories of one-time on-site work happen (control installation, geodetic adjustment, reconnaissance + coordination).
Realistic cost differential: 30-80% above steady-state for the first cycle, depending on what's already in place. Worth quoting first vs subsequent transparently rather than averaging.
Five things to lock in at first capture for cheap subsequent cycles: permanent control, flight plan documentation, pipeline version, deliverable format, independent reference data.
For some project types — environmental, civil, stockpile, flood — the first capture is the most valuable single deliverable because it establishes the baseline that every future cycle compares against. The overhead is the investment.
Tell us the site, the expected cycle frequency, and the foundational data you need locked in. We'll structure the first capture to deliver the permanent setup AND the project deliverable — and quote subsequent cycles transparently against the steady-state rate.
How first-capture overhead fits into the four pricing levers and the typical bands by project tier.
The most concrete example of why the first capture matters most — once piles are placed, the ground beneath is buried forever.