Lesson 5: Mapping the process (fishbone, SIPOC, value stream maps, the X-Y diagram, and FMEA)

This lesson opens the Measure phase, and with it, the second half of the Yellow Belt Body of Knowledge. Before a process can be measured meaningfully, it has to be defined clearly, on paper, in a way the whole team agrees on. This lesson covers four tools used to do exactly that, plus a fifth tool used to anticipate where a process is likely to fail before it does.

Cause and effect diagrams (fishbone or Ishikawa diagrams)

A cause and effect diagram, also called a fishbone diagram for its shape or an Ishikawa diagram after its creator Kaoru Ishikawa, is a structured brainstorming tool for organizing possible causes of a specific effect, such as a defect or a delay. The effect is written at the head of the diagram, and possible causes branch off a central spine, typically grouped into standard categories such as Man (people), Machine, Method, Material, Measurement, and Mother Nature (environment), sometimes shortened to the "6 Ms." The tool's value is in forcing a team past the first, most obvious cause that comes to mind and into a genuinely wide search across every category before jumping to conclusions.

SIPOC

SIPOC is a high-level process map that stands for Suppliers, Inputs, Process, Outputs, Customers. It is typically the very first mapping tool used on a new project, built at a high level, five to seven process steps at most, to establish a shared, agreed-upon boundary for what the project is actually looking at before anyone drills into detail. Suppliers provide inputs, inputs feed the process, the process produces outputs, and outputs go to customers. Building a SIPOC as a team exercise early in a project frequently surfaces disagreements about where the process even starts and ends, disagreements that are far cheaper to resolve in a one-hour SIPOC session than after weeks of data collection on the wrong boundary.

Value Stream Mapping

A Value Stream Map, or VSM, is a more detailed map than a SIPOC, showing both the flow of material and the flow of information through a process, step by step, with the actual time each step takes. Every step is classified as value-added, something the customer would knowingly pay for, or non-value-added, everything else, which is where the seven wastes from Lesson 4 typically get identified in concrete, measured form rather than as abstract categories. A current-state VSM documents the process as it actually runs today, including all its waste; a future-state VSM then depicts the target process once that waste is removed, and the gap between the two becomes the team's improvement roadmap.

The X-Y diagram

An X-Y diagram, also called a cause and effect matrix, is a structured way to prioritize a long list of potential input variables, the X's, against the outputs, the Y's, that matter to the customer. Each X is rated for how strongly it is believed to affect each Y, and the ratings are combined into a single score per X. The result is a ranked list showing which handful of inputs are most likely worth investigating first with real data, turning a brainstormed list of twenty or thirty possible causes into the small number a team can realistically chase down in the Analyze phase.

Failure Modes and Effects Analysis (FMEA)

FMEA is a structured tool for anticipating how a process or product could fail before it actually does, rather than waiting to discover failure modes after the fact. For each step in a process, the team identifies the ways it could fail, the potential effect of that failure, and its potential causes, then rates each failure mode on three factors: Severity, how serious the effect would be; Occurrence, how likely the cause is to happen; and Detection, how likely the failure is to be caught before it reaches the customer. Multiplying these three ratings together produces a Risk Priority Number (RPN) for each failure mode, which the team uses to prioritize where to focus preventive action, the same way Pareto analysis prioritizes which existing defects to chase first. A high RPN does not always mean the most severe failure; a moderately severe failure that happens often and is hard to detect can outrank a catastrophic failure that almost never happens and is always caught.