Nothing Changed Overnight: How Safety Controls Slowly Drift Towards Failure

A safety control can be effective when it is first introduced. The procedure is current. Workers understand what is expected. Equipment is maintained. Supervisors are checking compliance. Actions are followed up. Everyone is paying attention.

Six months later, the control may still exist on paper—but is it still working in practice?

Perhaps an inspection is being completed less thoroughly. A temporary workaround has become normal. A procedure no longer reflects how the task is actually performed. An overdue maintenance issue is being managed informally. Or workers have gradually adapted the process because the original control makes the job harder to complete.

Nothing has necessarily changed overnight.

Instead, the organisation has experienced safety drift: the gradual movement away from the conditions, behaviours and safeguards that made a control effective in the first place. And because the change can happen slowly, it can be difficult to see.

Controls don't remain effective simply because they exist

When organisations identify a hazard and implement controls, there can be a tendency to treat the job as complete. Workplaces change. Equipment ages. People move roles. Contractors come and go. Production pressures fluctuate. Procedures evolve. New technology is introduced. Tasks begin to be performed differently.

A control that was suitable when it was introduced may become less effective as the environment around it changes.

This is one reason WHS risk management does not end when controls are implemented. Safe Work Australia identifies reviewing control measures as a core part of the risk management process and advises organisations to regularly check that controls are continuing to work as planned.

Where a control is no longer effective, or workplace conditions change, it may need to be modified or replaced.

What does safety drift look like?

Safety drift is rarely dramatic. More often, it appears as a series of small deviations that individually seem insignificant.

For example:

• A pre-start inspection is still completed, but workers begin selecting the same responses each day without checking every item.

• A machine guard is removed for maintenance and starts being left off for short periods because replacing it slows production.

• A temporary exclusion zone remains in place for months and becomes part of normal operations.

• A procedure requires two people to perform a task, but experienced workers begin completing it alone.

• A corrective action is closed after a toolbox talk, even though the underlying behaviour continues.

• A critical control verification is completed, but repeated failures are looked at individually rather than as a pattern.

• A risk assessment remains unchanged even though the equipment, personnel or method of work has changed.

No single event necessarily triggers an incident. That can actually reinforce the new way of working. If a shortcut is taken repeatedly without consequence, people can gradually begin to view it as acceptable. The gap between work as designed and work as actually performed gets wider. Eventually, the control still appears to exist—but its reliability may have significantly weakened.

Why safety drift can be difficult to detect

Many organisations already collect large amounts of safety information through inspections, incident and near-miss reporting, hazard management, corrective actions, audits and control verification. The challenge is that these activities are often viewed as separate transactions rather than as connected indicators of how the safety system is performing over time.

An inspection is completed and closed, an action is assigned and completed, a near miss is investigated, and a failed verification is corrected. Individually, each process may appear to be working as intended. However, safety drift often becomes visible only when these records are considered together and over a longer period.

One failed control verification may be an isolated issue, but repeated failures of the same control across multiple sites may indicate a broader problem. Similarly, one overdue maintenance action may not seem significant, while recurring maintenance findings involving the same piece of critical equipment should attract greater attention.

The same applies to how work is performed. One worker deviating from a procedure may be a behavioural issue, but an entire team consistently performing the task differently may suggest that the procedure itself no longer reflects operational reality.

Individual records can tell you what happened. Looking at the pattern across those records can tell you whether the system itself is beginning to drift.

Warning signs that a control may be weakening

Organisations can look for a number of indicators that existing controls are no longer functioning as intended.

• Recurring inspection findings: If inspections repeatedly identify the same issue, completing another corrective action may not be enough. Repeated findings can indicate that the response is treating the symptom rather than addressing why the control continues to fail.

• Repeat incidents and near misses: Minor incidents and near misses can provide early warning that existing controls are being bypassed, misunderstood or are simply ineffective. Patterns involving the same activity, equipment, location or hazard deserve closer attention—even if nobody has been seriously injured.

• Corrective actions that solve the immediate problem only: An action can be completed without meaningfully improving the control. For example, retraining an employee may close an action quickly. But if workers repeatedly make the same mistake, the organisation may need to look deeper at task design, supervision, equipment, workload or the control itself.

• Increasing reliance on administrative controls: Over time, stronger controls can sometimes be supplemented—or quietly replaced—with procedures, signage, reminders and additional training. Administrative measures have an important role, but organisations should be alert when responsibility for managing a hazard gradually shifts from a physical or engineered safeguard towards workers simply being expected to remember more.

• Informal workarounds: Workers often adapt tasks to deal with practical operational problems. Those adaptations can provide valuable information. If employees regularly bypass, modify or work around a control, asking why may reveal that the control is difficult to use, poorly designed or no longer suited to the way the task is performed.

• Changes that haven't triggered a risk review: New equipment, new contractors, altered production methods, changed staffing levels, different materials or modified work environments can all affect the effectiveness of existing controls. Safe Work Australia specifically identifies workplace change as a reason controls may need to be reviewed.

Critical controls require greater visibility

Safety drift becomes particularly important when the control involved is relied upon to prevent a fatality or other serious event.

A critical control is not simply another item on a checklist. It is a control whose failure could materially increase the likelihood or consequence of a serious unwanted event. That means organisations need confidence not only that the control exists, but that it is present, effective and functioning when it is needed.

Periodic reviews alone may not provide that confidence. Critical controls may require defined performance standards, clear ownership and regular field verification.

For example, an organisation managing mobile plant interaction risks may rely on controls such as physical segregation, exclusion zones, communication systems and operator competency.

The important question is not:

“Do we have these controls?”

It is:

“How do we know these controls are working today?”

This is the shift from documenting controls to actively managing their effectiveness.

Use everyday safety activity to detect drift earlier

Preventing safety drift does not necessarily mean creating more safety processes. Often, organisations already have the information they need. The opportunity is to connect it.

• Inspection findings can show where a control is repeatedly failing.

• Incident and near-miss records can identify circumstances where controls were absent or ineffective.

• Risk assessments can provide the original context for why the control exists.

• Actions can show what has been done in response to failures.

• Critical control verification can provide evidence of whether high-consequence controls are performing to an agreed standard.

• Dashboards and trend reporting can then help identify deterioration that may not be obvious from individual records.

The aim is to move from: Find problem → create action → close action

towards: Identify weakness → understand pattern → strengthen control → verify effectiveness

That final step matters. Closing an action does not necessarily mean the risk is controlled.

How myosh can help organisations identify control drift

A connected safety management system can make it easier to identify when controls are weakening over time.

Rather than treating risk assessments, inspections, incidents, actions and control verification as isolated activities, myosh allows organisations to bring these processes together within a broader HSEQ environment.

With Critical Control Management, organisations can define critical risks and controls, establish performance standards and record ongoing control verification. myosh verification records provide a digital trail of control oversight, while inspections can be used to perform field-based checks of control effectiveness.

Smart Inspections can take this further by using inspection responses to trigger predetermined actions, notifications and escalation workflows when particular conditions are identified. Inspection data can also be analysed over time through reporting and trend charts rather than viewed only as individual completed checklists.

Where an issue requires intervention, Action Management can assign accountability, due dates and traceable corrective actions linked back to the records that generated them.

Digital Bowtie and Critical Control Management functionality can also help organisations maintain the connection between major unwanted events, threats, controls, verification and corrective responses instead of leaving that information in a static risk document.

Together, these processes can give safety teams a clearer picture of not just what controls should exist, but how those controls are performing over time.

Don't wait for failure to prove a control has drifted

One of the challenges with safety drift is that an organisation can operate for a long time with a weakening control and experience no serious consequence. That absence of harm can create confidence.

But it does not necessarily mean the control is effective. The better question is whether the organisation can demonstrate that the controls relied upon to manage its most significant risks are still functioning as intended.

Safe Work Australia advises organisations not to wait until something goes wrong before reviewing controls; control measures need to be maintained and reviewed as part of an ongoing risk management process.

For safety leaders, that means looking beyond whether procedures exist, inspections are completed and actions are closed.

Look for the gradual changes, the repetition, workarounds and looking at the controls that technically exist but no longer perform as originally intended. Because serious incidents do not always begin with one major failure. Sometimes they begin with a series of small changes that became normal enough that nobody noticed.

‍