Chapter 03
Fixed, mobile, 360, and drone
How to choose between the four main capture modalities for each workflow and combine them so the project gets coverage without paying twice for the same evidence.
01
Fixed cameras and where they earn their keep
Fixed cameras give continuous coverage at named work fronts. They are strongest on progress verification, gate logs, and incident readiness, because the value of the source increases with the time-aligned record it produces. A tower crane camera capturing the structural frame for fourteen months is irreplaceable; the same view shot weekly with a phone is not. Fixed cameras are weak when the work front moves quickly — a fitout floor that turns over every six weeks, an MEP riser that races up the building, a façade that closes off the view as it climbs. Plan fixed coverage where the front is stable for at least a quarter, and accept that some packages will need a different modality. The retention budget for fixed cameras compounds quickly, so the plan should specify which streams are retained at full resolution, which at downsampled, and for how long.
02
360 walks and route discipline through scaffold churn
360 walks bridge fixed and mobile. A walker carrying a 360 head along a defined route, at a defined cadence, produces a record that can be revisited later from any angle. 360 is strongest in fitout, MEP, and quality verification, where the value is in being able to inspect from any direction after the fact, often months later when a defect surfaces. The discipline is the route. A 360 walk without a fixed route is harder to compare across cycles, because the alignment between captures degrades and the as-built layer cannot be stitched. On a fitout floor with shifting scaffold, the route should follow the permanent reference points — column lines, lift cores, stair cores — rather than the temporary routes scaffold creates. The walker re-finds the route each cycle even if the floor looks different. A fortnightly cadence works for active fitout; monthly is enough once the floor is closing out.
03
Drone surveys, infrastructure cadence, and the project envelope
Drone or aerial captures cover what no other modality can — the structure as a whole, the earthworks, the roof, the surrounding environment. Drone is most valuable on large structural and infrastructure projects, on linear works such as roads and rail, and on projects with significant external scope such as data centres or energy. The cadence is usually monthly on buildings, fortnightly during earthworks, and weekly during peak structural lifts. On a linear infrastructure project, drone flights are tied to chainage rather than calendar — every five hundred metres of completed alignment captured to a consistent altitude and overlap. The combination of drone with fixed and 360 produces a record that holds up at every scale, from a single fitting to the project envelope. Drone alone is rarely enough; the survey is a wide reference layer that other modalities anchor to.
04
Mobile capture, helmet cameras, and confined-space inspections
Mobile capture — phone, tablet, helmet, or wearable — picks up what fixed and 360 miss. It is opportunistic, lower in average integrity, and most useful as a complement rather than a substitute. Helmet cameras come into their own on confined-space inspections, plant interiors, lift shafts, and tunnel headings, where no fixed mount can give the inspector view. The trade-off is integrity: mobile capture has shakier provenance, weaker time alignment, and higher tagging burden, so the workflows that depend on it should be limited to the ones that genuinely need a roving view. Helmet capture during a permit-controlled lift, recorded against the permit number, is high-value evidence; the same helmet running ambient through a working day produces a tape no one will retrieve. Specify the windows when helmet capture is on and off, or it becomes noise.
05
Combining the four without paying twice
The art is the combination. A monthly drone flight, two fixed cameras at the structural elevations, fortnightly 360 walks on active fitout, and helmet capture during named permit operations is a coverage stack that closes most building workflows. The redundancy is deliberate where it matters — a drone flight and a fixed elevation view both cover the frame, but at different scales — and absent where it does not. The cost trap is to add modalities because they are available rather than because they close a workflow. Each new source on the plan should carry the workflow it serves, the cadence it runs at, and the retention window it needs. A source without those three lines is one to question at the next revision.
Practice
01. Build a four-column matrix for your project: workflows down the left, modalities across the top. Mark each cell with the cadence and integrity grade. Identify any workflow with no primary source and any source with no workflow.
Look for: A finished matrix typically shows fixed cameras as primary for progress and gate, 360 as primary for quality and as-built, drone as primary for envelope and external, and mobile as supplementary across all four. Gaps usually appear as a workflow such as MEP commissioning or confined-space safety with no dedicated modality, and as a fixed camera placed for visual completeness with no workflow against its name.
02. Pick one workflow that is currently served by fixed capture only. Propose how a 360 or drone overlay would change the evidence quality, and estimate the marginal cost as a percentage of the existing capture spend on that workflow.
Look for: A typical answer takes progress verification on a fitout package, currently fixed-only, and proposes a fortnightly 360 walk that adds inspection-from-any-angle and as-built coverage. The marginal cost runs at ten to twenty per cent of the workflow spend, justified by the reduction in defects-cycle retrieval time and the as-built archive value at handover.
03. Specify the retention budget for one fixed camera over a thirty-month project. State the resolution, the storage tier, and the workflow that justifies the retention window.
Look for: A defensible specification might run twelve months at full resolution on a hot tier for active claims and progress, eighteen months at downsampled resolution on a cold tier for warranty correlation, and an indefinite hold on tagged clips associated with named delay events or incidents. The justification ties each tier to the workflow window it serves.
Checkpoint
For your current project, which modality is most under-used relative to the workflows in scope, and what would you cut to fund it?
Recommended reading