Distribution Center Route Risk: A 168-Site Safety Case

Distribution-center safety is often treated as a local operational problem, although the exposure usually belongs to the whole network. A route change, a crowded dock, a tired driver, a rushed load check and a weak handover can appear in different cities while carrying the same risk signature.

During Andreza Araújo's tenure at PepsiCo South America Foods, the documented outcome was a 50% reduction in accident ratio in six months, across a region that included seven countries, 30 factories and 168 distribution centers. This article does not repeat the broader 180-day follow-up rhythm. It studies one practical mechanism inside that work: making distribution route and dock exposure visible before the lagging metric confirmed the change.

The thesis is simple enough to test in any network. Distribution centers do not become safer because headquarters asks for more reports. They become safer when route risk, loading risk, pedestrian movement, contractor interfaces and driver fatigue are reviewed as connected exposure, with line ownership close enough to change tomorrow's work.

For an EHS manager responsible for several sites, the trap is to compare distribution centers only by accident rate. That view rewards quiet sites, punishes transparent sites and hides the operation where risk is rising but nobody has been hurt yet.

Initial scenario

The starting condition was a large food operation with different countries, dense logistics activity and many handoffs between factory, warehouse, transport provider and customer delivery. Each location could explain its own constraints, but the network needed a language that allowed leaders to compare exposure without flattening local reality.

In distribution, many serious events are preceded by signals that look ordinary until someone connects them. A near collision between forklift and pedestrian, a route assigned after a late production release, a driver waiting too long at a gate, a load secured under time pressure and a dock blocked by temporary storage may sit in separate logs. When leaders read them separately, the pattern stays invisible.

Andreza's documented executive background matters here because the PepsiCo role covered factories and distribution centers together. That scope prevented a common distortion in safety management, where manufacturing receives structured attention while logistics is treated as a contractor or transport issue. The exposure still belongs to the business decision, even when a third party is involved.

Decision

The decision was to move distribution risk from scattered local reports into a leadership rhythm that could be challenged across sites. This did not mean imposing a single checklist on every operation. It meant defining the few exposure questions that every site had to answer with evidence.

The first question was whether the site could show where people, vehicles and materials conflicted during normal work. The second was whether route changes, delivery pressure and contractor behavior were entering the safety discussion before an incident. The third was whether supervisors had enough authority to stop a weak loading or dispatch condition without being treated as the cause of delay.

This logic is consistent with Andreza Araújo's book Safety Culture: From Theory to Practice, where culture is treated as repeated operating behavior rather than declared intent. In a distribution center, culture is visible in who challenges a rushed load, who accepts a blocked pedestrian path, who escalates a repeated carrier problem and who treats a silent route as safe without verification.

Execution

The execution started by separating accident history from exposure visibility. Accident history showed where harm had already appeared. Exposure visibility showed where leadership needed to intervene before harm appeared. That distinction gave the safety team a way to discuss quiet but fragile sites without waiting for a recordable injury.

Supervisors and logistics leaders had to identify a short set of daily risk signals. The useful signals were not decorative dashboard numbers. They were conditions that changed work, such as late dispatch, dock congestion, reversing conflicts, missed load restraint checks, driver wait time, temporary pedestrian rerouting and repeated contractor nonconformance.

EHS did not own those signals alone. If EHS owned them alone, the process would become another inspection program. Operations had to own the work sequence, transport had to own carrier escalation, supervisors had to own the immediate field decision and senior leaders had to remove the structural blockers that made unsafe shortcuts attractive.

The adjacent dock safety visibility case follows the same principle at a narrower site level. The difference here is scale. A 168-site distribution network needs comparable questions, but it also needs enough local judgment to avoid turning risk review into a paperwork contest.

Measured result

The measured business outcome from the wider PepsiCo South America Foods work was a 50% reduction in accident ratio in six months, according to Andreza Araújo's documented executive track record. The distribution lesson should be read as one contributing mechanism inside that broader result, not as a claim that route-risk mapping alone produced the entire reduction.

That distinction matters because case studies become dangerous when they are turned into recipes. A plant with weak dock design, a mine with long-haul fatigue exposure and a retailer with urban delivery pressure will not copy the same controls. They can copy the discipline of making exposure visible, assigning decisions to the right owner and testing whether the field condition changed.

The companion article on the accident-ratio metric explains how the lagging result should be interpreted. In this case, the early indicators were more useful for daily leadership: repeated route deviations, unresolved dock conflicts, open carrier escalations, serious-risk action aging and field evidence of corrected pedestrian or vehicle flow.

Before and after indicators

The table is intentionally operational because distribution risk is operational. Senior leaders do not need poetic language to act. They need to see which repeated conditions are being normalized across sites while the monthly accident rate still looks acceptable.

Generalizable lessons

The first lesson is that logistics risk must be managed as part of the safety system, not as an external transport problem. When the business decides production sequence, dispatch timing, customer promise and staffing, it also shapes the risk that drivers, loaders, pedestrians and contractors absorb.

The second lesson is that quiet sites deserve scrutiny. James Reason's work on latent failures helps explain why a site can look clean on lagging indicators while tolerating weak conditions that have not yet aligned into an event. Silence is not proof of control, especially when reporting quality varies across locations.

The third lesson is that supervisor authority has to be explicit. If a supervisor can stop a loading condition only when the EHS manager is present, the system has not delegated safety authority. It has delegated blame after the fact, which is why decision rights should be tested during normal pressure, not only during campaigns.

The fourth lesson is that network leaders need one review that connects distribution exposure to executive decision. A monthly executive meeting that only receives injury rates will miss the weak signal. A useful executive critical-risk review asks which route, dock or contractor signals require resources, escalation or changed operating rules.

What to apply in your operation

Start with the last 30 days of distribution activity and choose five sites with different risk profiles. For each site, compare accident history with live exposure indicators: dock congestion, reversing movements, pedestrian deviations, load restraint defects, late dispatch, carrier recurrence and driver wait time. The contrast will show whether your quiet sites are controlled or merely underreported.

Then assign ownership by decision, not by department. EHS can define the method and challenge evidence, but operations must own work sequence, logistics must own carrier escalation, maintenance must own physical controls and supervisors must own the immediate stop or release decision. A risk register that names EHS as owner for every distribution exposure is already telling you that ownership is weak.

Finally, review the pattern weekly for six weeks before changing the dashboard. Leaders need to learn what good evidence looks like before the metric becomes formal. Otherwise, sites will learn to feed the indicator rather than expose risk.

When to ask for outside help

Outside support becomes useful when several sites disagree about what counts as distribution risk, when carrier issues repeat without escalation, or when the dashboard says performance is improving while supervisors describe pressure to release weak conditions. Those contradictions usually point to leadership design, not to a lack of awareness.

The PepsiCo case shows why distribution-center safety deserves executive attention. The visible result was a 50% accident-ratio reduction in six months, but the transferable lesson is sharper than the number. A network changes when leaders stop waiting for harm to validate risk and start treating daily exposure as evidence that deserves action.