The 168-Distribution-Center Safety Case: Giving Local Leaders Clear Route-Risk Decisions

A loading bay can look orderly while a driver, forklift operator and supervisor are each working from a different picture of the risk. That gap becomes harder to see when one safety standard is copied across a large network and local leaders are expected to make the right call without clear authority.

Andreza Araujo's experience at PepsiCo South America involved seven countries, 30 factories and 168 distribution centers. The useful lesson from that scope is not that one checklist can control every route. It is that a network becomes safer when local leaders know which decisions belong to them, which conditions require escalation and which controls must be visible before movement begins.

This is a case about governance at the point of work. It does not claim a measured reduction in injuries for the 168-center network, because that result is not part of the documented evidence available here. Its measured contribution is different because it shows how a distributed operation can define a common decision language without erasing local route conditions.

Initial scenario: one network, many route exposures

Distribution centers share familiar hazards, yet the exposure changes with the physical and operating context. A reversing movement beside a pedestrian door is not the same decision as a trailer exchange in a congested yard, even though both may appear under the broad label of vehicle interaction.

The same variation appears in contractor interfaces, dispatch pressure, weather, shift timing, dock design, maintenance status and the quality of communication between warehouse and transport teams. A central safety function can describe the expected control, but it cannot stand at every gate when a driver arrives with a different trailer, a barrier is out of position or a route changes during the shift.

That is where a common checklist begins to fail. It tells people what a good condition looks like, but it often leaves unanswered who can stop the move, who can change the route, who must approve an exception and how quickly the decision has to reach the next level.

The case matters because scale amplifies ambiguity. In a single site, an experienced supervisor may compensate for an unclear rule through personal judgment. Across 168 distribution centers, that informal compensation produces 168 variations of what leaders believe they are allowed to accept.

The decision: standardize the questions, not every local answer

The central decision was to separate the network's non-negotiable control expectations from the local judgment required to apply them. That distinction sounds simple, although it changes how a regional safety system is designed.

A central standard should state the exposure that must be controlled, the minimum evidence that the control is functioning and the conditions that require a pause or escalation. A local leader should then be able to decide how the route is arranged, how the workforce is briefed and which physical constraint must be corrected before the movement continues.

James Reason's work on latent failures helps explain why this separation matters. An incident is rarely created by a single visible act. Weak design, unclear responsibility, production pressure and missing feedback can align until the final error appears to be the whole story. Route-risk governance has to expose those upstream conditions before the vehicle starts moving.

That logic also fits the Headline Podcast's editorial focus. Leadership is not only the message delivered at the start of a shift. It is the permission structure that determines whether a local supervisor can challenge a route, delay a dispatch and ask for a different control without being treated as the cause of lost output.

Execution: build a route-risk decision chain

The practical work begins by translating the network standard into a decision chain that local leaders can use under pressure. The chain should fit the operation's real handoffs rather than reproduce the language of a policy manual.

The first question is whether the planned movement matches the approved route and traffic arrangement. If it does not, the supervisor needs authority to pause the task while the change is assessed. The second question is whether the critical controls are present and functioning, including segregation, visibility, communication, vehicle restraint and the condition of the loading area. A signed inspection is not enough when the field condition has changed.

The third question is who owns the unresolved exposure. EHS may support the assessment, but the operational owner should not disappear behind the safety department when the decision concerns dispatch, maintenance, warehouse flow or contractor coordination. The fourth question is what evidence closes the decision. A verbal assurance can start a conversation, but the route, barrier or work practice has to be observable before the movement resumes.

This chain is deliberately more demanding than a checklist because it tests decision quality. It asks whether a control can still perform its function under the conditions that actually exist, which is the point at which many routine inspections become too comfortable.

Measured result: governance coverage became the observable outcome

The documented scope gives the case a clear measured boundary. The operating model had to serve seven countries, 30 factories and 168 distribution centers. Those figures matter because they define the problem of scale, not because they prove an injury outcome.

That last row is important. A mature case study distinguishes what was measured from what sounds plausible. A network may improve decision clarity without an immediate movement in lagging indicators, because exposure, reporting quality and operating conditions influence those numbers. Leaders should therefore inspect the decision chain itself before using a later metric as proof of success.

For local managers, governance coverage can be tested through a small sample of real movements. Ask whether the supervisor knew the stop condition, whether the control owner was visible, whether exceptions had an expiry and whether the close-out evidence showed a changed condition. The sample does not replace injury data. It makes the operating mechanism inspectable.

What the case teaches about local ownership

Lesson 1. A network standard is useful when it clarifies the control purpose and the escalation threshold. It becomes cosmetic when it asks local leaders to copy a form without giving them authority to change the work.

Second, Route risk is a leadership issue because the most consequential decisions often happen before the vehicle moves. Dispatch timing, yard layout, contractor coordination and maintenance release can either strengthen or weaken the control environment.

Third, Local variation should be visible, not treated as evidence of poor discipline. A different route may be the right control for a different site, provided its owner, conditions and verification method are clear.

Fourth, EHS creates more value when it improves the quality of operating decisions instead of becoming the permanent owner of every decision. The line organization has to carry the authority that matches the exposure.

Fifth, Scale requires a common vocabulary, but it also requires a disciplined way to say, "This site is different, and here is the evidence." Without that sentence, local reality is either ignored or turned into an undocumented exception.

What to apply in a distribution network today

Choose one recurring movement that crosses warehouse, transport and contractor boundaries, then map the decisions that must be made before the movement, during a changed condition and after a pause rather than beginning with the entire network.

Then assign four owners. One person owns route design, another owns the physical control, another owns the dispatch or work authorization and a named leader owns escalation when the planned condition is no longer available. If one role appears in every box, the system is probably using EHS as a substitute for operational ownership.

Next, ask supervisors to describe the stop condition in their own words. Their answers should converge on the exposure and the escalation threshold, while still allowing local details to remain local. When the answers diverge, investigate whether the standard is unclear or whether the sites genuinely face different conditions.

Finally, review the evidence in the field. A route map, a barrier, a signal, a contractor handoff or a maintenance release should be observable. If the proof exists only in the system, the organization has measured record completion rather than control health.

Use the hierarchy of controls as a decision lens when a route depends too heavily on reminders or personal vigilance. Pair that review with the warehouse walking-surface inspection routine when the physical environment is part of the exposure. For contractor interfaces, the contractor mobilization safety plan offers a useful adjacent control question. When a change has to be held, compare the decision with the temporary change hold-point model before the route is reopened.

A large network does not become safer because every site uses the same form. It becomes safer when every local leader can identify the exposure, exercise the right authority and show why the movement was allowed to continue.

Conclusion

The 168-distribution-center case is valuable because it resists a common shortcut. Central standards do not remove local judgment, and local judgment does not excuse unclear controls. The operating model has to do both jobs at once.

When leaders define the decisions that belong at the point of work, route risk becomes easier to see and harder to pass upward as vague concern. That is the practical contribution of a distributed safety system. It gives local leaders enough structure to act and enough authority to prevent a routine movement from becoming an accepted exposure.