How 250+ Projects Turned Fatigue Signals Into Work Design Decisions

Fatigue signals become useful only when leaders treat them as work design evidence, not as a request for better coping. Across 25+ years in multinational EHS roles and more than 250 cultural transformation projects, Andreza Araujo has seen the same pattern repeat. The crew is not failing because people are weak. The crew is reacting to schedules, task sequencing, supervision gaps, and recovery debt that the system created.

This case study is a composite of repeated field patterns from Andreza Araujo's portfolio in 30+ countries. The thesis is direct. When fatigue moves from a personal complaint into a management decision, leaders stop asking who needs to try harder and start asking which part of the operation made safe performance less likely.

Key Takeaways

• Treat fatigue as work design evidence, because overtime, recovery, and task timing shape risk before anyone calls it a wellbeing problem.
• The strongest fix is not a reminder to sleep better. It is a faster change in roster, task mix, handover quality, or supervision.
• ISO 45003:2021 and the WHO and ILO 2022 mental health at work brief place psychosocial risk inside the operating system, not outside it.
• Supervisors need a same-shift response rule, because a fatigue signal that waits for a monthly meeting is already late.
• A quiet crew can hide tiredness, so leaders need to verify schedules, recovery windows, and critical task exposure instead of trusting tone alone.

Initial Scenario

The starting point looked like a wellbeing issue. The site had overtime, night work, a few close calls, and a crew that said they were coping. Managers responded with sleep hygiene reminders, short talks about rest, and the usual advice to be careful on the next shift. The dashboard improved only cosmetically because the roster, the handover, and the priority stack stayed the same.

That is the trap Andreza Araujo describes in Far Beyond Zero. Fragile mental health makes physical safety fragile, because fatigue, excessive load, and weak support reduce attention and slow the decision that should stop the task. A tired worker is not a moral failure. A tired worker is a predictable output of a schedule that keeps borrowing from recovery.

In a 30+ country portfolio, the pattern appeared often enough to become obvious. The most exposed crews were usually the ones asked to absorb schedule pressure, task changes, and a weak handover at the same time. The organization saw the symptom late, then tried to correct the symptom in the same meeting that produced it.

Decision

The decision was to move fatigue out of the wellness lane and into the operating lane. Every fatigue signal had to answer three questions. What in the schedule created it? What task would become unsafe if the signal was ignored? What can change before the next shift starts?

That shift matters because ISO 45003:2021 treats psychosocial risk as part of occupational health and safety management, and the WHO and ILO 2022 brief on mental health at work points leaders toward prevention, support, and work design, not just individual resilience. If the problem lives in the work system, the response must also live there.

Andreza Araujo's experience across more than 250 cultural transformation projects points to a simple rule. Leaders do not improve safety by asking the worker to absorb more pressure with the same design. They improve safety by changing the conditions that make pressure predictable.

Execution

Workload became visible

The first move was to name workload in operational terms. The team mapped overtime, consecutive night shifts, short recovery windows, high consequence tasks after midnight, long handover chains, and the points where the crew was most likely to lose attention. The point was not to label people as stressed. The point was to show when the job itself was asking more than the system could safely absorb.

That map changed the conversation. A supervisor could no longer say only that the crew was tired. The supervisor had to show which task, which hour, and which staffing pattern made fatigue foreseeable. Once the exposure was visible, the operation could stop treating tiredness as a private matter.

The response window moved

The second move was to shorten the response window. A fatigue signal no longer waited for the next weekly review. It triggered a same-shift decision with an owner, an interim control, and a date for follow-up. If the issue involved a high risk task, the job paused until the exposure was changed or accepted by the correct authority.

This is where the article How to Run a Post-Overtime Fatigue Debrief in 12 Minutes becomes useful, because a fatigue conversation only works when the first response is structured. If the first answer is vague, the worker learns that reporting fatigue only creates noise.

The handover stopped hiding fatigue

The third move was to turn handover into a control point. Outgoing crews had to record changed conditions, unresolved fatigue signals, and any task that had already been stretched by time pressure. Incoming crews had to confirm what changed, what control was weaker, and what decision could not wait until morning.

That is a direct application of Andreza Araujo's safety culture logic in Safety Culture: From Theory to Practice. Culture shows up in repeated decisions, especially when the shift is late and the simplest answer is to keep going. A useful handover reveals the risk instead of hiding it inside production notes.

Supervisors got one rule

The fourth move was to give supervisors one rule that they could remember under pressure. If fatigue touches a high consequence task, the supervisor must change the task, add verification, reassign the work, or escalate before release. That rule sounds simple because it should be simple. The site does not need more language. It needs a repeatable decision.

This also changed what supervisors measured. They stopped counting only whether the crew finished the shift. They started asking whether the work was still fit for the people assigned to it. That is a better control question, because it focuses on the task that actually creates exposure.

Measured Result

The defensible result from the portfolio is not a single universal percentage, because the projects span sectors, sites, and operating models. The consistent result is narrower and more useful. Leaders closed the distance between a fatigue signal and a management decision. The crew stopped waiting for a monthly meeting to discuss a problem that was already visible in the same shift.

In the stronger sites, the conversation changed from who is not coping to what in the system made fatigue predictable. That is the real measure. A fatigue program works when it changes the question, because the new question forces leaders upstream to roster design, handover quality, and task sequencing.

During Andreza Araujo's PepsiCo South America tenure, where the accident ratio fell 50% in 6 months, the lesson was the same. Results improved when leaders changed the operating rhythm, not when they asked people to be more careful inside the same weak flow. Fatigue follows that rule as well.

Generalizable Lessons

• Fatigue is not a resilience test. It is a control problem that shows up in the schedule, the handover, and the task design.
• A quiet site can still be a tired site, because people often stop reporting when they think nothing changes.
• Awareness training helps only when the operation also changes staffing, recovery, and escalation rules.

The deeper thesis is that work design creates the conditions under which mental strain becomes a safety issue. That is why the article Mental Health Triage Drift: 6 Failures That Delay Safe Work Decisions belongs in the same conversation. If the organization treats distress as an isolated HR event, it will miss the way pressure moves through the shift.

The Illusion of Compliance is useful here because a clean schedule on paper does not prove that the crew has enough recovery in the real week. The file can look orderly while the human system has already been overdrawn.

What to Apply in Your Operation

Start with one crew that combines overtime, night work, and high consequence tasks. Review the last 30 days of rosters, handovers, near misses, and overtime approvals. Then ask which task becomes unsafe first when sleep debt rises. That is the task that should drive the trigger.

Next, define one trigger that supervisors can use without debate. Examples include consecutive nights, missed breaks, repeated micro errors, a rushed restart, or a high consequence task after the circadian low point. Connect that trigger to a clear response, because a trigger with no response only teaches people to ignore the rule.

Then link the response to work design. If the issue is predictable overtime, change the roster. If the issue is a weak handover, change the handover format. If the issue is a task that should never be done after a certain hour, move it or add verification. The companion guide Work Ability Explained: 4 Dimensions Behind Safe Return is useful when the team needs to separate capacity, health, and task fit.

Finally, make operations own the schedule and EHS own the trigger. HR and occupational health should support the process, but they should not be asked to carry a problem that lives in shift design. That is how the control stays close to the work.

FAQ

What makes fatigue a safety issue?

Fatigue becomes a safety issue when tiredness changes attention, judgment, handover quality, or the ability to stop a high consequence task. At that point the problem is no longer only personal comfort. It is exposure.

Why is a wellness message not enough?

Wellness messaging can help people recover, but it does not change the roster, the deadline, or the task sequence that created the fatigue signal. If the work design stays the same, the signal usually returns.

Who should own fatigue control?

Operations should own the schedule and staffing decisions that create fatigue, while EHS should define the trigger and test whether the response protects the task. HR and occupational health should support the process, especially when recovery or fitness for work needs review.

What is the first sign leaders should watch?

Look for repeated micro errors, missed breaks, short recovery windows, and handovers that only describe production status. Those are early signs that fatigue is already affecting the system.

How fast should a fatigue signal be answered?

It should be answered in the same shift whenever the signal touches a high consequence task. A fatigue signal that waits for the next meeting is already too old to protect the work.

Recommendation

Do not ask the crew to carry fatigue better. Ask the operation to design it out where possible, contain it where necessary, and escalate it before a high consequence task starts. That is the practical difference between a wellbeing slogan and a control system.

If your leadership team wants to translate this case into its own schedule, handover, and escalation rules, follow Headline Podcast for the leadership side of the conversation and use Andreza Araujo's work as the diagnostic lens.