Process Map

Overview

A process map (flowchart) graphically depicts a workflow — steps, decisions, inputs, and outputs are arranged in logical order. While a SIPOC only shows the frame, the process map describes the actual sequence and branches.

Activity (rectangle): A single work step — e.g. "Measure part", "Enter order". Activities change the state of the process.

Decision (diamond): A branching point with yes/no or multi-answer outcomes. At least two paths leave a decision.

Start and end (ovals): Define the beginning and end of the process. A clean map has exactly one start and at least one end.

Arrows (flow direction): Arrows give the order. They should be readable without crossings — otherwise re-sort the layout.

Swim lanes: Optional horizontal or vertical lanes showing which role, department, or system owns a step. Makes hand-offs and interfaces immediately visible.

Process maps exist at various levels of detail — from a roughly sketched high-level map to a detailed standard operating procedure. For SIPOC context, "rough" is enough; for standardization, "detailed" is needed.

Value classification (VA / BNVA / NVA): Each step can be classified as value-adding (VA), business non-value-adding (BNVA), or non-value-adding (NVA). VA steps create direct customer value. BNVA steps are operationally necessary but provide no customer value. NVA steps are waste and candidates for elimination. Click the badge to cycle through types.

Input type: Process Parameter (x) vs. Noise Factor (n): Each input can be classified as a process parameter (x) or noise factor (n). Process parameters are controllable inputs — machine settings, recipes, set points. Noise factors are uncontrollable or hard-to-control inputs — ambient temperature, batch variation, operator variability. The distinction helps with downstream DOE planning and root cause analysis.

Loops (feedback loops): Use the loop icon in a step's header to define a feedback loop back to an earlier step. The condition describes when the loop triggers (e.g. "If measurement out of tolerance"). Additional loop steps describe actions that only execute on the feedback path. A visual bracket on the right edge shows the loop span.

Approach

• Define start and end first — what triggers the process, what is the end result?
• Add the main steps as activities (rectangles) — no decisions yet, only flow.
• Add decisions as diamonds — typically at points where "it depends".
• Connect with arrows — keep flow direction consistent (top → bottom or left → right).
• If useful, add swim lanes for responsibilities.
• Review the map with the people who do the work — clearly separate as-is from to-be.

A first map shows the as-is state. Only afterwards do you design the to-be state — keep both versions for before/after comparisons.

Pitfalls

As-is mixed with to-be: Drawing while "improving" loses the true as-is capture. First cleanly document what actually happens — including the ugly workarounds. Improvement comes later.

Hand-offs invisible: Weak spots often sit at the interfaces between departments. Without swim lanes they stay invisible.

Too much detail: A map with 200 steps nobody reads. For DMAIC analysis the essential steps are enough; detailed standards belong in a separate SOP.

Only one person interviewed: Every person sees the process from their angle. A complete map emerges only when all participants have been heard.

No update after changes: Processes change — maps often do not. An outdated map is worse than none, because it carries false assumptions forward.

Loops without exit: Loops without an exit path are a diagram bug — they often point to a forgotten decision or a missing termination criterion.

Examples

This module ships with the following example datasets — load any of them in the app with a single click.

Available in the following cycles

• DMAIC: Measure
• DMADV: Measure
• 8D: D2 — Describe the Problem