How to Run a Start-of-Shift Hazard Scan in 15 Minutes

Every shift begins with a decision, whether supervisors name it or not. A start-of-shift hazard scan is the moment when the crew decides whether today still matches the plan from yesterday or whether weather, people, equipment, or simultaneous work has changed the risk. Across 25+ years of executive EHS work and more than 250 cultural transformation projects, Andreza Araujo has seen the same pattern repeat: people rarely miss the hazard itself, they miss the change that made the hazard active.

James Reason's work on latent failure explains why that matters. A line, machine, route, or task can look ordinary until a maintenance delay, a tired crew, a new contractor, or a hot afternoon shifts the system into a different state. The scan is not a paper exercise. It is the first control that decides whether work starts, slows, or stops.

What you need before starting

Before the crew gathers, the supervisor needs the task plan, active permits, weather, staffing, equipment status, last-shift notes, and one question that matters more than the rest: what changed since the plan was written?

A toolbox talk shares information. An inspection confirms condition. A hazard scan decides whether today's exposure still matches yesterday's assumptions. That difference is why the scan belongs before the job starts, not after the crew is already standing in the work area.

Bird and Heinrich still matter here because precursor events are the signal, not the injury count that comes later. If the crew treats the scan as a formality, the operation keeps its paperwork and loses its early warning.

Step 1: Define the work window

Start by naming the exact job window, area, and sequence. If the task includes lifting, hot work, mobile equipment, open edges, or line-of-fire exposure, say so in plain language. The scan fails when it describes the work in broad terms because broad language hides simultaneous operations.

Verify the window by repeating the job in one sentence without using the word usual. If the task cannot be described cleanly, the crew may still be carrying an old version of the work in its head. The common error is to talk about productivity before the exposure has been named.

Step 2: Map active exposures

Walk the route or section and note what can hurt people now. Focus on energy sources, traffic, edges, weather, chemicals, noise, congestion, and any line-of-fire condition that puts the body between force and consequence. Not every hazard matters equally, so the scan should rank the ones that can injure someone in the next few minutes.

Verify the ranking by asking which hazard would create the fastest harm if the crew did nothing differently. If the answer is vague, the scan has become a list instead of a decision tool. The common error is to let the crew name ten hazards and still leave the worst one unranked.

Step 3: Check control drift

Compare the live field to the barrier you expected. Is the guard still in place, is the barricade intact, is the eyewash reachable, is the ventilation on, is the bypass still off, is the access route still clear? The point is not to inspect everything again. The point is to detect drift in the controls that this job depends on.

Verify by touching, seeing, or testing the specific control, not by trusting last week's sign-off. A signed checklist can be true and still be stale. The common error is to assume that a completed form means the barrier still works in the field.

Step 4: Read the people signal

The crew can be competent and still unsafe if the shift has the wrong mix of fatigue, language, new joiners, role overlap, or pressure to recover time. Ask who is new, who is returning from absence, who is covering two roles, and who will make the stop decision when the task gets noisy.

Verify by having the person most exposed to the hazard repeat the critical step in their own words. If they hesitate, the supervisor has found a training gap or a process gap before the work starts. The common error is to talk only to the most senior worker and assume that the rest of the crew heard the same thing.

Step 5: Test change points

This is where James Reason becomes practical. A change point is any condition that makes a routine job unstable: weather, schedule compression, maintenance deferral, unusual material, contractor overlap, shift handover, or a temporary field change. If the task depends on a control that was altered overnight, the scan should treat that as a risk event, not as a note in the margin.

Verify by asking what changed since the last safe run and what changed again after the handover. If the crew cannot answer, the work is probably running on memory instead of current conditions. The common error is to let a temporary change stay temporary until someone is hurt by it.

Step 6: Decide the control action

Every scan should end with one of three actions: proceed, modify, or stop. That decision must belong to one person, and that person must be visible to the crew. If everyone agrees but nobody owns the call, the work starts on social pressure instead of control.

Verify the decision by naming the owner before the crew disperses. That name should be the same person who can delay the job, change the sequence, or ask for another control without asking permission from the loudest voice on site. The common error is to use the scan to collect opinions while avoiding a decision.

Step 7: Say the stop criteria aloud

The scan only protects people when stop criteria are spoken in plain language. If the guard is missing, if the route is blocked, if the weather changes, if the permit is incomplete, or if the control room cannot support the task, the crew should hear the trigger before the job starts. Silence creates room for wishful thinking.

Verify by having the supervisor state the first stop condition out loud. When the crew can repeat it without improvising, the rule has moved from memory to shared practice. The common error is to keep the criteria in the supervisor's head and expect the field to guess correctly later.

Step 8: Verify at the field edge

The final step is physical confirmation. Stand where the work will happen, not where the forms sit, because that is where the field tells the truth. The scan should close with a short field note, the name of the control owner, and the next review time if the work has been modified.

Verify that the note could support a later review without relying on memory. If a different supervisor read it at the end of the shift, they should know exactly what changed, who owns the next action, and why the job was allowed to proceed. The common error is to end the scan at the trailer table and call that field verification.

Final checklist

• The work window is named in one sentence.
• The active exposures are ranked by immediate harm.
• The live controls were checked at the field edge.
• The people signal was read, not assumed.
• The change points were named before work started.
• One visible owner made the proceed, modify, or stop call.

If one of those six items is missing, the scan is incomplete, even if the form is signed and the crew is already moving. The point of the routine is not to sound careful. The point is to prevent a normal start from becoming an avoidable exposure.

FAQ

Is a hazard scan the same as a JSA?

No. A JSA explains how the task should be done, while a hazard scan checks whether today's conditions still support the task at all. The scan comes first because it answers the start question before the crew spends time on task detail.

Who should lead the first-hour scan?

The supervisor who can change the plan should lead it, because the scan only works when the person speaking can also act on what the field shows. A delegate can assist, but ownership should remain visible and undisputed.

What should trigger a stop?

Any missing control, unplanned overlap, unstable weather, unclear ownership, or unknown change point can justify a stop. The useful rule is simple: if the crew cannot explain how the exposure is controlled today, the job should not start as planned.

What if the crew says there are no issues?

That answer is not enough on its own. The supervisor still needs to check the route, the controls, the people, and the change points, because silence can mean confidence, but it can also mean habit, pressure, or incomplete awareness.

Why does this matter for safety culture?

It matters because culture shows up in the first decision, not in the poster on the wall. When leaders reward a fast start more than a correct start, people learn to hide uncertainty. When leaders reward a clear stop call, the crew learns to surface risk early.