FlowMapper AI — an online VSM tool for value stream mapping

Why this tool exists

Value Stream Mapping is one of the most effective Lean tools — provided someone actually sits down, collects reliable data on the shop floor or in the office, and draws the map so it can actually yield conclusions. In practice, most VSM maps I've seen over years of work as a Lean Six Sigma Master Black Belt are made in Excel or on paper, are rarely updated, and even more rarely does anyone recalculate PCE or Takt Time on the fly as the data changes.

I wanted a tool that reverses the order of work: a person collects data where they know the subject best — on the gemba, talking to the operator or the person running the process — while the tedious calculating, drawing and keeping the map consistent is handled by the tool. Hence FlowMapper AI: an input sheet with hints (what to measure and how), automatic drawing with classic VSM symbols, and metrics calculated instantly, so the team's discussion starts from facts, not from drawing boxes.

I separately solved a problem that classic VSM often overlooks: processes where one step is created from several (sometimes dozens of) parallel inputs — raw materials, sub-assemblies, information from different sources. In FlowMapper AI, such a step can be expanded into a tree of inputs, as deep as needed, and you can "drill into" any level to see a separate map and metrics calculated only for that branch.

This tool is free for simple processes. The PRO version unlocks full scale — any number of steps and an unlimited, multi-level component tree — for companies that want to map genuinely complex processes.

What the tool does

• Two data modes — manufacturing (C/T, C/O, Uptime, cells) and services/office (turnaround time, switch-over time, availability, case queue), each with its own terminology and hints for every column.
• Automatically drawn VSM map — classic symbols: a process box with data (C/T, C/O, Uptime), an inventory triangle, a "push" arrow, an electronic-information zigzag from the control cell, supplier and customer.
• Continuous timeline (Lead Time ladder) — one step chart beneath the map: a high step = days of waiting, a low step = seconds of real processing, with a summary total-Lead-Time cell.
• Recursive components and raw materials — any step can have its own parallel inputs (sub-assemblies, raw materials), each with its own inputs — no depth limit. Above 5 inputs, the map automatically collapses them into a single marker to keep it readable.
• Drill-down — clicking any component shows a separate map and metrics calculated only for that branch, with breadcrumbs to go back up to any level.
• AI insight cards — automatic observations about the bottleneck, the largest buffer and low process efficiency, with guiding "5 Whys"-style questions.
• "What if" simulator — sliders for inventory, changeover time and turnaround time for a selected step, with a live before/after map and metrics.
• Print and PDF export — directly from the browser, with a watermark and footer.

How the metrics are calculated — full description

All the formulas below work identically in manufacturing and service mode — only the column names change, not the math.

Takt Time

This is the pace at which the customer "takes" one unit or one case. If a process step's cycle time (C/T) is longer than Takt Time, that step can't keep up with demand — the tool automatically flags it as a bottleneck (a red outline on the map, a yellow Kaizen marker).

Inventory / queue in days

The raw number of units or cases waiting in the buffer doesn't say much on its own — only once divided by daily demand does it show how many days of work that inventory will cover. That's exactly the value drawn as a triangle on the map and as a high step on the timeline.

Total inventory / Lead Time

This is the key difference from an Excel sheet: when a step has an expanded tree of sub-assemblies and raw materials, their inventory doesn't disappear from the calculation — it adds up to the total Lead Time together with the main line. That's exactly what shows up on the timeline as a highlighted "Lead Time" cell at the end of the ladder.

Value-added (VA) time

This is the total time of real work on the product or case — also calculated recursively through the whole tree, not only along the main line.

PCE (Process Cycle Efficiency)

The classic process-efficiency metric: what percent of the total lead time is real, customer-paid-for work. Values below 5% are very common in practice — they mean the product or case spends most of its time waiting, not being processed. The tool flags such a result in red in the metrics panel.

How nested levels (components and raw materials) are calculated

Every main-line step can have its own parallel inputs — components, sub-assemblies or raw materials that are actually produced in separate cells and only come together at that one step (e.g. a chair's side frame welded from five parts, each with its own cycle time and buffer). The calculation rules are the same at every level:

• Each component has its own C/T, C/O, Uptime and inventory, exactly like a regular step.
• A component whose C/T exceeds Takt Time is flagged as a bottleneck — same as a main-line step.
• A component can have its own components (a raw material feeding a sub-assembly, a sub-assembly feeding another sub-assembly) — with no depth limit.
• The top-level metrics (Takt Time, Lead Time, VA Time, PCE) always sum the entire tree — every level, not just the main line.
• Drill-down: clicking any component takes you to a separate map that shows that component as a single step with its own input tree above it — and the metrics at the top recalculate only for that branch. Breadcrumbs above the map let you return to any level above, up to the full main map.
• Collapsing with many suppliers: when a step has more than 5 parallel inputs (typical with dozens of raw-material suppliers), the map automatically shows a single collapsed "▤ N inputs" marker instead of cluttering the view — clicking it expands the full list.

Symbols on the map — classic VSM language

Open the tool

Below is the working application — you can enter your process data right away. Free version: 1 main map, max. 3 steps, no input components.

License: free and PRO

The free version is not a trial — there's no time limit. It only limits the scale of the process. When your process is bigger, or you need input components, I'm happy to tailor a PRO license to your situation — get in touch, a short conversation is enough.

The online tool above is always the free version — for a quick look at how it works. You run the full PRO version from the downloaded file, locally in the browser — deliberately, because a shop floor doesn't always have access to an external network, and mapping a process shouldn't depend on that.

How to get started (4 steps)