Ask most HSE managers what a person on board roster is for and they will usually say headcount: who is on the asset right now, so a muster can be called and emergency response can start. That's correct, but it's only half the job. The other half, the part that gets far less attention until an audit or a drill exposes it, is whether everyone on that roster is actually competent and authorised for the role they're filling today.
What a POB management system actually needs to verify
A POB management system records who is physically on an asset at a given moment, for muster and safe manning purposes. On its own, that's a headcount. For it to support emergency response and manning compliance properly, it also has to confirm that each person's certifications, inductions, and task-specific competencies are current and match the role they're performing that shift.
That second part is where most gaps actually sit. A roster can be perfectly accurate as a count and still tell you almost nothing about whether the people on it can do what an emergency scenario, or a routine permit, requires of them.
Safe manning obligations differ by asset type
It helps to separate what legislation actually requires from what good practice adds on top, because the two get blurred in a lot of vendor material.
For marine vessels, safe manning sits with the flag state. A Safe Manning Document, issued under flag state requirements informed by STCW, sets the minimum number and grades of crew needed to operate the vessel safely, including watchkeeping and emergency duties. For fixed and floating offshore installations, there isn't an equivalent single certificate. Instead, the duty holder's safety case has to demonstrate that the installation can be safely evacuated, escaped, and rescued, which under PFEER 1995 means the emergency response arrangements, including temporary refuge capacity and evacuation means, have to be sized and tested against the number of people who could realistically be on board.
That distinction matters in practice. A DSV or a flotel can have marine crew operating under one competency framework and installation personnel operating under another, counted on the same muster list but assured against different rulebooks. Reconciling that is its own problem and worth treating separately from the core POB and competency question this piece is about.
Why competency records are the weak link
In our experience designing workforce and Control of Work systems for operators, the POB count itself is rarely where the real uncertainty lives. It's the competency data feeding it. Training certificates sit with the provider. Contractor-specific competencies sit with the contractor's own system. Site-specific inductions and task authorisations sit with the client. None of these systems are required to talk to each other, and on assets running a high turnover of contractors and visiting personnel, reconciling all three against a live roster by hand is slow enough that it tends to happen at the gangway, once, rather than continuously.
The result is a roster that's confident about who is on board and much less confident about whether those people are currently qualified for what they're doing right now, as opposed to what they were qualified for when they were last inducted.
Reconciling the roster and the competency record before anyone counts as safely on board
Safe manning, in a workable operational sense, depends on three things working together: a real-time, accurate POB record; verified and in-date competency data for every person on that record; and a process that checks the two against each other before someone is allowed onto the asset, or counted as validly there.
That reconciliation is where access control, training records, and the POB system either connect or don't. We've written before about why a person on board tracker matters offshore, and the same logic applies to competency: a tracker that only answers "are they here" is solving half the problem a muster actually needs solved.
What an inaccurate roster actually costs you
Consider a hypothetical muster drill where the headcount in the assembly area matches the expected total exactly, but nobody present currently holds a valid certificate for operating survival craft, because the two people who did have rotated off three days earlier and the roster's competency field was never updated. The count is right. The readiness is not. Nothing about that scenario shows up if the only thing being checked is whether the numbers add up.
The same gap shows up in less dramatic ways day to day. A permit gets issued to a work party where the roster confirms everyone is on board but the system doing the checking has no visibility of whether the banksman named on the lift plan is actually one of them, or whether their certification has lapsed since the last job. That's an operational assurance problem as much as a permit one, which is why we'd point readers toward operational assurance: certifications and KPIs for the broader framework this sits inside.
What to check in a POB and competency management system
Rather than taking a vendor's word for it, there are specific things worth asking them to demonstrate:
- Show how an expired or soon-to-expire certificate is flagged against a named individual before they board, not after.
- Show how a competency gap for a specific task surfaces at the point of permit issue or muster, rather than only at induction.
- Show how contractor and third-party competency records are reconciled against the live roster, including where those records originate from a separate training provider system.
- Show how the roster updates when a mid-rotation crew change happens, and whether the outgoing and incoming personnel are both correctly reflected during the handover window.
Many of these same questions come up when assessing contractor competency specifically, which we've covered in more depth in contractor permit management: closing the competency gap on critical equipment.
Where this leaves manning compliance
A roster that's accurate as a count but disconnected from competency data will pass a casual glance and fail a proper test. Getting the two reconciled isn't primarily a software problem, it's a question of which process decides, before someone steps onto the asset, that their presence and their qualification are both verified together. If you're reviewing how your current POB and competency data sit relative to each other, it's usually worth mapping where the two actually connect today before deciding what needs to change.