Chapter 02
Incident evidence preservation
How to preserve the captured base around an incident so the record survives default rotation cycles and the dispute window that follows.
01
The first hour and the second hour
In the first hour after a significant incident, the priority is care of the injured, control of the area, and notification. That is the human work and it takes precedence over everything else. In the second hour, while the area is still secured and before the day moves on, the priority shifts to preservation. Someone with named authority places a hold on the captured records around the time and location of the incident. The hold covers the views before, during, and after the event, the gate logs for the day, the personnel scans for the package, any 360 walks that pass through the area in the preceding fortnight, and any drone surveys taken in the relevant period. The hold prevents default rotation from quietly deleting records the inquiry will later need. The work is small at the time and decisive later. The team that knows how to do this in the second hour saves itself months of reconstruction. The team that does not will be assembling the record under the pressure of a regulator deadline, and will likely find that the relevant footage was overwritten on day thirty-one.
02
What the preservation set actually contains
The preservation set is broader than people expect on the day. The captured view of the work front itself. The views of the access route to the work front, because struck-by exposures and lifting incidents often have causes a hundred metres upstream. The gate logs for that day and the day before. The personnel scans matching the package and the activity. The 360 walks taken in the days before, which often show the condition of edge protection, exclusion zones, or housekeeping that contributed. The method statement, the permit, the lift plan, and any deviation notes. The shift records and the toolbox attendance. The weather record if the activity was weather-sensitive. Each of these may be requested by the regulator, the insurer, the client, or a later inquiry. Preserving them now, while the picture is fresh and the access is easy, is much cheaper than reconstructing them three months later from emails and recollection. A preservation policy that lists these explicitly removes the on-the-day judgement call.
03
The preservation policy and the named authority
The team should not be making the preservation decision under the pressure of the incident. The policy already exists, written down, agreed with the project director and the HSE lead, and circulated. It names the trigger events that require a hold (any RIDDOR-reportable event, any near-miss with major potential, any regulator visit, any insurer-notifiable incident). It names the authority — typically the HSE lead, with the project director as backup — who can place the hold, and the duration before review. It names what gets preserved, in the form of a checklist matched to the activity type. It names how the hold is released when the inquiry concludes. The policy is short, perhaps two pages. It belongs alongside the capture plan and the emergency response plan. The discipline is to know it before the day, so the second hour can be acted on rather than negotiated. The platform supports the policy by allowing the hold to be applied with a single action, with the held records flagged in the retention system so they survive the default rotation.
04
The conversation with custody and what good looks like
A preserved record is only useful if its custody is defensible. The chain of custody for the held records — who placed the hold, when, against which incident reference, and who has accessed the records since — is part of the record itself. The platform writes that audit trail without manual effort. When a regulator asks where the footage came from, when it was preserved, and who has touched it, the answer is a single export rather than an apology. What good looks like is a project where the second hour after an incident is procedural rather than improvised, where the preservation set is wider than the inquiry will need, and where the chain of custody is complete enough that the recording can be entered into evidence without challenge. The discipline costs minutes; the absence costs months.
Practice
01. Write a one-page preservation policy for your current project. Cover trigger events, the named authority, the preservation checklist by activity, the default hold duration, and the release process.
Look for: A two-column page with the five sections above. The named authority should be a role and a person, with a backup. The checklist should differentiate between lifting, working at height, confined space, hot works, and access incidents.
02. For a hypothetical struck-by incident in your busiest delivery yard, list every captured record you would preserve and the reason for each. Include records you might not initially think of.
Look for: A list of at least eight records: gate logs, delivery manifests, fixed views of the yard, views of the approach road, banksman positions if recorded, the lift plan, the relevant permit, weather, attendance, and any 360 walks of the area that fortnight.
Checkpoint
Does your project have a written preservation policy, and do you know who can place a hold within the second hour today?
Recommended reading