Back to Insights

Method statement, permit, risk assessment: is the high-risk work actually ready?

Aven-AI InsightsHealth & Safety7 min read
Two construction workers in hard hats working at height on metal scaffolding.

Citable answer: High-risk work is not ready just because it appears in the programme. The method statement, permit and risk assessment must be in place before the work starts. Aven-AI's predictive look-ahead flags missing readiness items early enough for humans to fix them.

There is a gap between "the paperwork exists" and "the work is safe to start," and a lot of incidents fall straight into it. A risk assessment is on file. A method statement was issued at tender. A permit to work has a number on it. Everything looks complete on the register — and yet the controls described on paper are not the controls in place at the workface. The question that matters before a high-risk activity begins is not do we have the documents? It is is this specific work, today, actually ready?

What the documents are, and what they are for

The two core documents usually travel together as RAMS — a Risk Assessment and a Method Statement. They are separate but linked. As Notify Technology describe it, the risk assessment identifies the hazards of a task and how serious they are, and the method statement then draws together the conclusions of those risk assessments and sets out, step by step, how the job will be done and how each hazard is controlled (Notify Technology).

A permit to work is a further layer for the highest-risk activities — a formal authorisation confirming that specific precautions are in place before that particular task starts, in that particular location, for a defined window. As SafetyCulture note, method statements are specifically expected for high-risk activities such as working at height, excavations and demolition, and for activities requiring a permit to work (SafetyCulture).

What CDM 2015 actually expects

In the UK, the Construction (Design and Management) Regulations 2015 require a safe system of work to be developed for construction activities — delivered through the construction phase plan and, at task level, through RAMS. As HammerTech explain in their guide to RAMS in UK and Ireland construction, principal contractors and clients commonly require RAMS before granting site access and use them to judge a contractor's competence and commitment to safety (HammerTech). The HSE is clear that paperwork is a means, not an end: the duty is to plan, manage and monitor the work so it is carried out safely, not merely to hold the documents (HSE).

That last point is the one that gets lost. A signed RAMS file satisfies an administrator. It does not, by itself, satisfy the regulation. The regulation is about the work being controlled.

Where readiness actually breaks down

The documents can all be present and the work still not be ready. The common failures:

  • The RAMS is generic, not task-specific. A template method statement that could describe any excavation does not control this excavation, with its actual ground, services and access.
  • The risk assessment is stale. Conditions changed — weather, sequence, an adjacent trade — but the assessment still describes last month's site.
  • The permit references controls that are not in place. The permit assumes an exclusion zone, a competent person, a piece of plant — and one of them is not actually there when the crew arrives.
  • The chain is broken between documents. The method statement relies on a risk assessment that was never updated, or a permit issued against an earlier version of the method. Each document looks fine alone; together they no longer line up.
  • Competence and briefing are assumed. The people doing the work were never briefed on the method, so the controls on paper never reach the workface.

None of these show up as a missing tick on a register. They show up as a real hazard that the paperwork said was controlled.

A readiness check, before the crew starts

Treat readiness as a set of questions, not a filing status:

  1. Is the method statement specific to this task, location and conditions today?
  2. Does the risk assessment reflect the site as it actually is now?
  3. Is every control the permit relies on physically in place and verified?
  4. Do the three documents reference each other's current versions — or is one pointing at a superseded one?
  5. Have the people doing the work been briefed on the method and are they competent for it?

If any answer is "not sure," the work is not ready, regardless of what the register says.

Where a governed AI layer helps

This is a consistency-and-currency problem across documents that humans struggle to cross-check under time pressure. A system that reads the RAMS and the permit together can flag the mismatches a tired supervisor misses at 6am: a permit relying on a control the method statement never describes, a risk assessment older than the last significant change on site, a method statement that is generic where the task is specific. It does not approve the work — a competent person still makes that call. It makes sure the call is made on documents that actually agree with each other and with the site.

General health-and-safety information, not legal or professional safety advice. CDM duties and safe systems of work must be assessed by competent persons for the specific work and jurisdiction.

Sources & further reading

Ready to put this into practice?

See how Aven-AI flags notice, time-bar, programme and QHSE risks before the window closes.

Request early access