Safety Management & STOP

Most organizations reach a point where safety performance stops improving. The rules are written, the PPE is issued, incidents are investigated, and the numbers flatten — or move around randomly at a level nobody is happy with.

The reason is that rule-based safety addresses conditions, and by this stage most of the obvious conditions have been fixed. What remains is behaviour, and behaviour does not respond to another rule. The remaining incidents come from shortcuts taken under time pressure, habits formed because the safe method is inconvenient, and hazards nobody reported because reporting felt pointless or risky.

The clearest symptom is near-miss reporting. If your near-miss reports are few, that is not good news — near misses vastly outnumber incidents in any operation, so a low count means they are happening and not being reported. And a near miss unreported is a free warning discarded.

STOP — Safety Training Observation Program — is a behaviour-based approach built on structured observation. Trained observers watch work as it is actually performed, discuss what they see with the person doing it, and record the observation. The purpose is not enforcement; it is to surface the gap between how work is designed and how it is really done.

Done properly, this produces a specific and valuable insight: when someone consistently works unsafely, it is usually because the safe method is harder — the tool is awkward, the guard slows the job, the correct PPE is stored too far away, the procedure assumes time the schedule does not allow. That is a design finding, not a discipline finding, and it is actionable.

Done badly, observation becomes surveillance. Observers catch people out, records feed disciplinary processes, and within weeks people simply work safely when observed and revert afterwards. The programme then reports excellent observation scores and unchanged incident rates.

Everything therefore depends on how it is introduced: observations must be non-punitive and non-attributable, and observers must include peers rather than only supervisors.

Some organizations are still running safety systems built around OHSAS 18001, which has been withdrawn and superseded by ISO 45001. If that describes you, migration is overdue, and it is more than a renumbering exercise.

ISO 45001 differs in ways that require real work:

• Worker participation and consultation is a substantive requirement, not a committee that meets quarterly. This is the largest gap for most migrating organizations.
• Leadership obligations are explicit and cannot be delegated to a safety manager
• Context and interested parties must be understood, including contractors and visitors
• Risk and opportunity thinking replaces hazard identification alone
• It uses the Annex SL high-level structure, so it integrates cleanly with ISO 9001 and ISO 14001

That last point is the opportunity. If you hold ISO 9001 and ISO 14001, migrating to ISO 45001 is the moment to build a single integrated system rather than a third parallel one — and to apply document simplification while you are restructuring anyway.

We design the programme and train the observers, then support the first observation cycles — because the first few weeks establish whether people experience it as help or as surveillance, and that reputation is very hard to change later. We build the near-miss and hazard reporting route, and the feedback loop that makes reporting worthwhile: a report that produces a visible response generates more reports; one that disappears ends them.

Where migration is needed, we handle the move from OHSAS 18001 to ISO 45001 and integrate it with your existing management systems.

Avoid:

• Observation records feeding disciplinary processes — this destroys the programme quickly and permanently
• Observers drawn only from supervision, so it reads as management surveillance
• Treating consistent unsafe behaviour as a discipline issue when the safe method is harder
• Interpreting low near-miss reporting as good performance
• Reports that produce no visible response
• Still running withdrawn OHSAS 18001 and assuming it remains adequate
• Treating ISO 45001's worker participation requirement as a quarterly committee
• Building a third parallel system instead of integrating with ISO 9001 and ISO 14001
• Safety metrics based only on lagging indicators, which report history rather than predict risk

We bring 30+ years, 900+ organizations and a 100% first-time pass rate to this work.