Context
A published case study by Ferreira, Silva and Mesquita examines a manufacturer of domestic water heating equipment that ran a DMAIC project on one of its internal logistics processes. The paper presents a case study highlighting how a manufacturer of domestic water heating equipment has used the Six Sigma Define-Measure-Analyze-Improve-Control (DMAIC) methodology to improve one of its internal logistical processes, the replenishment of supermarket in production lines. This is the kind of internal, unglamorous process that logistics Green Belts are often handed: not a customer-facing delivery route, but the milk-run that keeps parts moving from stores to the production line inside the plant.
A milk-run is a fixed-route internal delivery system: a tugger or small vehicle follows a set path at set intervals, dropping components at line-side “supermarkets” so operators never run out of parts. When a milk-run route takes longer than planned, either the line risks starving for parts or the company has to add headcount and vehicles to compensate. That gap between planned and actual route time is what triggered this project.
The problem
The problem centred on uncertainty around the delivery time on these routes, uncertainty that created buffer stock and planning headaches for the lines it served. The paper describes how a manufacturer of domestic water heating equipment used the Six Sigma DMAIC methodology to improve one of its internal logistical processes, the replenishment of supermarket in production lines, moving it toward a world class quality level. In plain terms: three milk-run routes existed, and nobody could reliably say how long any of them would take on a given cycle. Some runs finished on time. Others ran long enough to threaten the line-side supply, and the company had no way to predict which days would be problem days.
How the belt scoped it
A Green Belt project has to be small enough to run part-time over a few months, which means the scope has to be bounded before any data collection starts. In this case the scope was one clearly defined process: the milk-run replenishment cycle serving the production lines, measured by route completion time. The project did not try to redesign the whole internal supply chain, renegotiate supplier contracts, or touch external freight. It picked one measurable output, route time, with a specific and painful symptom: a meaningful share of runs breaching the 30-minute target.
That scoping choice matters for anyone studying this as a template. A Green Belt charter should name the process, the boundary (start and end point of the route, in this case), the primary metric, and the pain point in numbers wherever possible, not a vague complaint about “slow logistics.”
Tools applied, phase by phase
Define
The team defined the process as the fixed-route milk-run cycle from stores to line-side supermarkets, with route completion time as the Y. A SIPOC-style map would typically sit here: driver, vehicle, route, stops, and the production lines as customers of the process.
Measure
The team collected route time data across cycles and routes to establish a baseline. Establishing a baseline distribution is exactly the step where a Green Belt would run a process capability study against the 30-minute target, and check measurement system reliability before trusting the numbers.
Analyze
With baseline data in hand, the analysis stage is where a Green Belt trainee would apply the statistical toolkit taught at this level: comparing route times across drivers, shifts or routes with t-tests or ANOVA to see whether differences were due to assignable causes or common variation, and using root cause tools such as fishbone diagrams and stratification to separate route design issues from driver behaviour or stop sequencing. The published account focuses on the outcome rather than every intermediate statistical test, but the structure of the project, moving from a variable baseline to a clearly diagnosed set of causes, follows this same logic.
Improve
Improvements targeted the sources of variation and excess time identified in Analyze, most likely route sequencing, stop consolidation and standardising how loads were staged. The result was a shift not just in average time but in the spread of times, which is the signature of a genuine variation-reduction fix rather than a one-off lucky day.
Control
To hold the gain, a Green Belt would put the new route standard on a control chart tracking route time by cycle, with limits tied to the 30-minute target, and a response plan for out-of-control points. That is the mechanism that prevents a project from drifting back to baseline six months after the belt moves on to something else.
What the result was
The documented outcome is concrete. The application of the Six Sigma methodology resulted in a reduction of those routes taking more than 30 minutes to be completed from 25 to 3 percent, a reduction in the coefficient of route time variability from 40 to 14 percent, and a reduction of the mean route time from 31 to 24 minutes. These results had a significant financial impact, allowing the elimination of one of the three existing routes without any negative impact in the supermarket replenishment process, leading to a drop in man hours and costs through the elimination of two milk runs.
That last point is worth pausing on. The financial payoff here did not come from a single dramatic fix. It came from squeezing variability down far enough that the company could safely consolidate three routes into fewer, which cut vehicles and labour hours without ever touching the customer-facing side of the business. That is a common shape for internal logistics Green Belt projects: the win is capacity freed up, not a defect count reduced.
What a trainee should take from it
| Lesson | Why it matters at Green Belt level |
|---|---|
| Scope to one measurable process | The project stayed inside one internal route system, not the whole supply chain, which kept it doable part-time. |
| Track variability, not just the average | The variability reduction, not the mean time alone, is what let the company safely cut a route. |
| Let the data justify structural change | Consolidating routes was only defensible once the data showed the remaining routes could absorb the load reliably. |
| Control plans protect the gain | Any route-time win needs an ongoing chart and response rule, or the old variation creeps back. |
For a trainee studying logistics applications of DMAIC, the value of this case is that it shows a Green Belt-scale project delivering a hard operational outcome, fewer vehicles and fewer man-hours, from what looks on the surface like a routine internal delivery problem. It is a reminder that the discipline of process mapping, capability analysis, root cause work and a control plan applies just as much to a forklift route inside a factory as it does to a customer-facing shipment.
