What Auditors Actually Look For in a Permit-to-Work Trail — And Where Most Sites Fail

An auditor pulls up a completed permit-to-work (PTW) record. The work finished three weeks ago. The permit is still showing as open.

Nobody on site did anything wrong, exactly. The isolation was done properly, the work was completed safely, the crew moved on. But the record doesn't reflect any of that — because the paperwork never caught up to the site.

This is the moment that generates most PTW audit findings. Not a missing signature. Not a forgotten checklist. A gap between what actually happened on site and what the record says happened.

If you manage PTW on a construction site in Singapore or Malaysia, this probably sounds familiar. And it's worth understanding why it keeps happening — because the fix isn't stricter supervision.

PTW problems are behavioral

It's tempting to treat PTW issues as a discipline problem — someone was careless, someone rushed, someone forgot to close it out. But look closely at almost any PTW finding and you'll notice the same root cause: the record-keeping step is physically and mentally separated from the verification step.

The permit gets signed in an office or site container, based on a check that happened — or should have happened — somewhere else, at some other time. The record and the reality are two different events, connected only by memory and paperwork.

That separation is where almost every PTW mistake originates. Three show up more than any others.

Mistake 1: Permits get signed before the condition is actually verified

On paper, PTW approval means: someone checked the site, confirmed it's safe for the specific work, and signed off. In practice, the signing often happens before the check — or based on a check that's no longer current.

A supervisor issues a permit for hot work at 8am based on a gas test taken the day before. By the time the crew starts at 9:30am, conditions may have shifted — ventilation changed, another trade moved into the zone, isolation was disturbed by unrelated work.

What this triggers in an audit:

Timestamp mismatches are one of the first things reviewed. If the permit issue time doesn't align with a verifiable, current condition check, it reads as a control failure — regardless of whether anything actually went wrong on site. Under Singapore's WSH framework and Malaysia's DOSH requirements, the expectation is that risk controls are verified as current at the point of work authorization, not assumed from an earlier check.

Downstream cost:

Beyond the audit finding itself, this is the exact gap that shows up in incident investigations when something does go wrong. A permit that was technically signed but not truly current offers no real protection — legally or operationally.

Mistake 2: Permits exist in three different places

Ask to see a completed permit trail and watch how many places someone has to check: a physical copy in the site file, a photo someone took and sent on WhatsApp, a scanned PDF sitting in an inbox.

None of these are wrong on their own. The problem is that no single one of them is the authoritative record — so nobody can produce one clean, time-stamped trail from issue to closure.

What this triggers in an audit:

Auditors aren't just checking that a permit exists. They're checking that it's traceable — one consistent record showing who verified, who approved, when, and under what conditions. Fragmented records force reconstruction after the fact, which almost always exposes gaps that a single system wouldn't have.

Downstream cost:

Reconstructing a trail after an audit request — or worse, after an incident — takes time your team doesn't have and rarely holds up as well as a record created in real time. It also puts individual supervisors in the position of personally vouching for a process, rather than the process speaking for itself.

Mistake 3: Closure is treated as optional

Work finishes. The crew is already on the next task. The permit — technically still active — never gets formally closed.

This is easy to overlook because nothing appears to be going wrong. Work got done safely. But an open permit for completed work raises a specific question during an audit: was there ever a defined closure check, or did the work just... stop?

What this triggers in an audit:

Closure is meant to confirm the site was returned to a safe state — isolations removed, barriers taken down, the area handed back. An open permit for finished work suggests that confirmation never formally happened, even if it informally did. That's a control gap on paper, even when the site itself is fine.

Downstream cost:

Open permits accumulate. On a busy site, it's common to find multiple "active" permits for work completed weeks earlier — which makes the entire PTW register harder to trust, audit, and rely on in an emergency.

What "good" actually looks like: a sequence, not a checklist

The common thread across all three mistakes is sequencing. PTW works when it follows a strict order, applied at the point of work — not reconstructed afterward:

Verify → Sign → Work → Close

• Verify the site condition immediately before authorization — not the night before, not from memory.
• Sign at that same point, so the approval reflects current conditions.
• Work proceeds under an active, current permit.
• Close the permit as soon as work ends, confirming the site is returned to a safe state.

None of this requires a new policy if not just closing the physical and time gap between checking and recording.

Where digital forms fit in

This is where most sites hit a practical wall: the sequence above is easy to describe and hard to enforce with paper forms, WhatsApp photos, and after-the-fact data entry. The verification happens on site. The record gets created somewhere else, sometime later. That gap is exactly what creates the three mistakes above.

At the core of the problem is the form itself. A paper permit — or a digital one that just replicates paper — is still a static document that gets filled in, carried around, and handed off between people and places. Every one of those handoffs is a place where the timing can slip, the copy can go missing, or the closure step can quietly get skipped.

GoBuid's digital forms are built to remove those handoffs entirely. The permit is filled in, verified, and signed directly on site, on the device already in the supervisor's pocket — not written up later from memory or a photo. Photos, GPS location, and timestamps attach automatically as part of filling in the form, so the record reflects the actual moment the check happened, not a reconstruction of it.

Because the form lives in one place from issue to closure, there's no version sitting in a noticeboard, another in someone's chat history, and a third in an inbox. It's one record, updated in real time, visible to whoever needs to see it — supervisor, safety manager, or auditor.

And because closure is just another step in the same form rather than a separate task someone has to remember, it's far harder for a permit to quietly stay "open" after the work is done.

None of this changes what a PTW process is supposed to do. It just makes the correct sequence — verify, sign, work, close — the only sequence that's actually easy to follow.

If you only fix one thing

Look at how much time passes between when a site condition is actually checked and when that check becomes a signed, recorded permit. If that gap is minutes, your PTW process is probably solid. If it's hours — or reconstructed at the end of the day — that's where your next audit finding is going to come from.

Want to see how a field-first ePTW workflow closes that gap in practice? [Talk to the GoBuid team →]