VSM Studio
How to use this course
Who it is for: plant, process and continuous-improvement engineers, supervisors and managers who need to map a value stream and lead the change. No prior lean training is assumed; experienced practitioners can skim module 1 and start at module 2.
What you need: a computer with Chrome or Edge (a phone works for viewing and small edits), VSM Studio open at builder.html, and — for the capstone — data from your own process: demand, shift pattern, and for each step its cycle time, changeover, uptime, operators, scrap and the inventory in front of it.
How each module works: learning objectives → the idea → step-by-step in the app → a hands-on exercise → questions to check yourself (answers fold open). Two practice maps ship with the app: Brackets (a 5-step metal line, in seconds) and the 3D print shop (7 steps with office work and unattended printers, in minutes). Load them from New map.
1 Lean and VSM foundations
You will be able to
- Explain what a value stream is and what a VSM shows that other charts don't.
- Define takt time, lead time, value-added time and the seven wastes.
- Name the four steps of value stream mapping.
The idea
A value stream is everything — value-adding or not — that turns raw material into a product in the customer's hands, for one product family. A VSM draws it on one page: material flow along the bottom, information flow along the top, and a timeline that compares how long parts wait with how long they are worked on. That comparison is usually the shock: days of waiting for minutes of work.
Lean asks five things of a value stream: produce at the customer's pace (takt), let work flow without stopping, pull where it can't flow, level the schedule, and keep improving. Waste shows up as overproduction, waiting, transport, over-processing, inventory, motion and defects — on a VSM most of it appears as inventory triangles and long ladder steps.
The method has four steps: 1 pick a product family, 2 map the current state, 3 design the future state, 4 plan and track the implementation. VSM Studio supports all four. The Guide covers the theory in more depth.
Check yourself
Demand is 460 parts/day and there are 2 shifts of 480 minutes with 20 minutes of breaks each. What is takt?
Available time = 2 × (480 − 20) = 920 min = 55,200 s. Takt = 55,200 ÷ 460 = 120 s per part.
Why map one product family rather than the whole plant?
Customers buy products, not departments. A family shares most downstream steps, so one map shows one flow you can actually redesign; a whole-plant map becomes unreadable.
Which waste does an inventory triangle make visible?
Inventory and waiting — and usually overproduction upstream that created it.
2 Getting around VSM Studio
You will be able to
- Find the palette, canvas, data panel, simulation dock and Assistant.
- Switch between Basic and Advanced, and change the language.
- Open, rename, copy and back up maps.
Step by step
- Open the builder. It starts on a blank map with four ways in: a form, the Assistant, dragging symbols, or a sample. Press ? Tour for the guided tour.
- Left: the symbol palette. Centre: the canvas — drag empty space to pan, scroll to zoom, F to fit. Right: the data panel with tabs (Data, Goals, KPIs, Plan, Mix). The ‹ › tabs on the canvas edges hide either side; click a section title to fold it.
- ⚙ Settings: Basic or Advanced, single features on or off, English / Español, start blank, and your own AI key.
- Save & share: rename, save a copy, version history, share link or file, print to A3/PDF, PNG/SVG, CSV import. Maps save automatically in the browser.
- On a phone, the bottom tabs switch between Map, Symbols, Data, Simulate and Assistant; turn the phone sideways (or press ⤢) for the side view.
Exercise
Open New map → Sample: current state. Rename it “Brackets — training”. Save a version called “Start”. Turn on Español in Settings, look around, and switch back. Hide the palette with ‹ and bring it back.
Check yourself
Where is your work stored, and how do you keep a backup?
In this browser. Back it up with Save & share → Save to a synced file (OneDrive, Google Drive or Dropbox folder) or Map file (.json); versions protect you from your own edits.
What does Basic hide?
Goals, KPIs, Costs, Mix, Operator balance, pull-system symbols and the less common process fields. Nothing is deleted — switching to Advanced shows it again.
3 Mapping the current state
You will be able to
- Collect the right data on the floor.
- Build a current-state map three ways: form, Assistant, by hand.
- Connect material and information flows and read the timeline.
Before you map: walk the floor
Walk the whole stream once, then walk it again from shipping upstream. Time at least five cycles per step with a stopwatch and count inventory where it sits — don't copy numbers from the ERP. Print the data sheet from the Guide.
Three ways to build it
- Form (fastest): New map → Start from a form. Enter demand, staffed minutes, supplier and shipping, then one row per step: time, batch, operators, uptime, changeover, scrap, rework and the queue in front of it. Build the map.
- Assistant: ✦ Assistant → Map my current process, and answer its questions — or paste a written description with a process table and it reads the numbers itself (with an AI key it understands free text).
- By hand: drag Customer (top right), Production control (top middle), Supplier (top left), process boxes left to right, inventory triangles between them. Select a symbol and drag its green handle onto the next to connect; pick the flow type in the palette (push, shipment, manual, electronic…). Right-click anything for colors, size, shape and flow route.
Then click empty canvas and set demand and working time; the takt box appears under the customer and the lead-time ladder builds itself under the map.
Exercise
Build this line with the form: demand 400/day, 450 staffed minutes, times in seconds. Steps: Cut 40 s (uptime 95%), Bend 55 s (92%), Weld 60 s (95%, scrap 2%), Pack 30 s. Queues before Bend 1,200, before Weld 1,600, before Pack 400; finished goods 600. Check: takt 67.5 s, lead time 9.5 days, processing 185 s.
Check yourself
Where does one process box end and the next begin?
Wherever flow stops and inventory piles up. Steps that pass parts one at a time without a queue belong in one box.
Why does the map read inventory in days?
Days = quantity ÷ daily demand — it turns a pile of parts into how long a part waits there, which is what lead time is made of.
4 Reading the numbers
You will be able to
- Interpret takt, lead time, processing time, PCE, rolled FPY, load and line capacity.
- Tell elapsed time from touch time and staffed days from calendar days.
- Use the issue markers and tooltips on the map.
The summary tiles
- Takt — the pace the customer sets. Load — what a step must make ÷ what it can make; over 85% there is no slack, over 100% it cannot keep up (the box turns red). Line capacity — good parts/day before the busiest step runs out.
- Lead time — days in inventory along the ladder. Processing time — the bottom steps. PCE — value-added ÷ lead time, often below 1%.
- Rolled FPY — product of every step's 100% − scrap − rework: the share that goes through right first time.
- In batch maps (elapsed time per batch), the totals separate queue days (staffed production days) from elapsed clock time and split it into touch and unattended time — they are never added into one number.
Hover the % load label on a box, a triangle or a kaizen burst: the tooltip explains the problem and what to do. The Load-vs-capacity chart shows every step at a glance.
Exercise
Open the 3D print shop sample. Find the constraint and explain why printing is at 98% load (hover it). Why is rolled FPY 84% when no single step is below 95%? How many days do parts wait, and how much of the 18 hours of processing is unattended?
Check yourself
Printing runs 6 printers at 90% uptime, 20 parts per build, one build per printer per day, 3% scrap. What can it make?
6 × 0.9 × 20 × 0.97 ≈ 104.8 good parts/day — against ~103 needed after downstream scrap, hence 98% load.
Why not add 4.2 queue days to 18 hours of processing to get "lead time"?
They are different clocks: queue days are counted in staffed production days at demand; processing is elapsed calendar time, much of it overnight. Adding them double-counts and mixes units.
5 Simulation and quick suggestions
You will be able to
- Run the value stream day by day and read what breaks.
- Test an improvement against the same run before changing anything.
- Use 💡 Suggestions to apply, test or mark improvements.
How the simulation works
Each day the customer orders the daily demand. Steps work during staffed time (unattended machines keep running), breakdowns follow each step's uptime, scrap is remade from the first step, and push queues hold work as long as the map says it waits — so the current state reproduces your map. FIFO lanes, supermarkets and continuous flow pass work on right away. A warm-up fills the line before the counted days start.
- Press ▶ Simulate, then Play (or Run instantly). Rings show each step working, waiting, blocked, broken down or off shift; numbered bubbles on the triangles show the queue; hover either for details.
- At the end read What the run shows: missed days, the constraint, growing queues, blocked steps, scrap. Flagged symbols get a red dashed outline with a tooltip.
- Under Try an improvement, press Test in simulation on an idea: the same demand and breakdowns run with and without the change.
- 💡 Suggestions lists the same ideas any time, with Test, Apply to map (Ctrl+Z undoes) and Mark with kaizen.
Exercise
On the print shop: run 20 days. How many days were late? Test “Add a second daily cycle at Printing” and “Replace push queues with FIFO lanes”. Which one fixes delivery, and which one cuts lead time? Mark both with kaizen bursts.
Check yourself
Why can a step at 98% load still miss days?
Average capacity covers demand, but breakdowns are random; with 2% slack any bad day becomes a late shipment.
Why do FIFO lanes cut lead time without adding capacity?
They cap how much work can wait between steps and pass it on immediately, so parts stop sitting in queues released once a day.
6 Designing the future state
You will be able to
- Answer the eight future-state questions.
- Build a desired state with the Assistant or by hand.
- Place FIFO lanes, supermarkets, the pacemaker and kaizen bursts.
The eight questions
- What is takt?
- Finished-goods supermarket or straight to shipping?
- Where can work flow continuously?
- Where do you need FIFO lanes or supermarket pull?
- Which single step is the pacemaker?
- How will you level the mix there?
- What increment of work (pitch) will you release?
- Which improvements are needed to make it work?
- With the Assistant: ✦ Assistant → Design the desired state. It analyses the open map, asks the eight questions with suggested answers, and builds “… — desired state” with FIFO lanes, the pacemaker, a leveling box, kaizen bursts and a before/after table.
- By hand: New map → Copy this map as future state, then replace push queues with FIFO lanes or supermarkets (right-click → Shape), add withdrawal and kanban flows, a leveling box before the pacemaker, and kaizen bursts where something must change.
Exercise
Design the print shop's desired state with the Assistant: make to order, FIFO lanes of one batch, pacemaker at Order & CAD review, level by pitch, keep all suggested improvements. Compare waiting days and rolled FPY before and after, then simulate the new map.
Check yourself
Why only one pacemaker?
One scheduling point keeps the whole stream in step; everything after it flows FIFO, everything before it is pulled to replace what was used.
Takt 4.5 min, batches of 20 — what is the pitch?
4.5 × 20 = 90 minutes: release one batch to the pacemaker every 90 minutes.
7 Goals, limits and costs
You will be able to
- Set goals and compare current, future and planned maps against them.
- Plan to a target within limits on people, machines, WIP and inventory value.
- Add costs and read cost per part and inventory value.
- On the Data tab, click empty canvas and add a labor rate, material cost and carrying cost; give each process a machine cost. Cost per good part and inventory value appear in the summary.
- On Goals, pick presets (Double output, Halve waiting, FPY 99%, Cost −10%) or add your own. The table compares this map, the future state and the plan.
- Set limits and press Suggest changes. The planner tries the cheapest fixes first — uptime, quick changeovers, restarting unattended machines, another shift — before buying machines, skips anything that breaks a limit, and names what is still short. It then simulates the plan and warns when “on paper” isn't enough.
- Build future state from this plan creates the map with kaizen bursts and a notes box.
Exercise
On Brackets: goal “Double output”. Plan with no limits, then with 6 extra people and 1 machine. What output can you reach within the limits, and which steps are still short? Add “Highest load ≤ 80%”, plan again and compare the simulation result.
Check yourself
Why does the planner aim for about 90% load, not 100%?
Variability: at 100% every breakdown is a late order. The built-in simulation shows the effect; 80–85% is safer.
8 Implementation plan and KPIs
You will be able to
- Turn kaizen bursts into owned, dated actions grouped into loops.
- Track progress on the map and in an A3 print.
- Choose and log KPIs against targets.
- On the future-state map open Plan and press Add the kaizen bursts. Each becomes an action, assigned to the pacemaker, upstream or supplier loop.
- Give each an owner, due date and status; overdue items turn red, and the burst on the map shows a status dot (✓ done, amber in progress, red blocked or late).
- Implement the pacemaker loop first, then work upstream. Review the plan weekly; export it as CSV or print the A3.
- On the current-state map open KPIs → Suggest KPIs: waiting days, WIP, yield, load at the constraint, changeover, raw stock, cost per part, on-time delivery, output. Targets come from the future state. Log actual values weekly; each KPI shows on target, improving or off target with a trend line.
Exercise
Create the plan for your print-shop desired state, assign owners and dates over the next 12 weeks, mark one action done and one blocked. Suggest KPIs on the current state and log a value for “Time in queues”.
Check yourself
Why start with the pacemaker loop?
It sets the pace for the whole stream and is closest to the customer, so its improvements show up in delivery first.
9 Advanced tools
You will be able to
- Drill from a plant-level map into a process with detail maps.
- Use the product-family matrix, EPEI and batch sizes.
- Balance operators against takt.
- Import data from CSV or an ERP export.
Detail maps
Right-click a process → Create detail map. The new map takes the steps before and after as supplier and customer, inherits demand and takt, and shows ↑ to go back; the parent box gets a ⤵ badge. Map the inside, then on the parent press Use these numbers to roll up capacity, pace, processing time, yield and operators.
Product mix and EPEI
On Mix, add each product with its demand and tick the steps it uses. Products sharing ~80% of steps form a family. EPEI = changeover time for every variant ÷ time left each day after running demand; the table gives the batch size per product. Short changeovers → short EPEI → small batches.
Operator balance
With Operator balance on, the Data tab shows each operator's work per part against takt and how many operators a balanced line needs (total work content ÷ takt).
Import
Save & share → Import CSV / ERP export. One row per step; columns are matched by name (work center, cycle time, setup, OEE/uptime, headcount, scrap, WIP…). Update the open map by step name, or draw a new map. Download the template to see the format.
Exercise
Create a detail map for Weld with three steps (fit-up 40 s, tack & weld 105 s at 92%, grind 35 s), roll the numbers up and see the parent's load change. Then add two products to the print shop (70/day and 30/day) and read the EPEI at printing.
10 Sharing and reporting
- Share link — the whole map packed into a link; it opens in VSM Studio wherever the app runs. Share map file — the .json by email or chat; they open it with Import.
- Save to a synced file — choose a file in a OneDrive, Google Drive or Dropbox folder; every change is written there while the tab is open (Chrome/Edge). Version history — named snapshots you can restore or open as a copy.
- Print / PDF (A3) — title block, the map, the summary and the implementation plan on one A3 landscape sheet. PNG / SVG for slides.
- Español in Settings switches the interface; with an AI key the Assistant answers in Spanish too.
Exercise
Print your print-shop desired state to PDF (A3), save a version “Ready for review”, and copy a share link.
11 Capstone: map your own value stream
- Pick the family — use Mix to build the product-family matrix and choose one family.
- Walk and collect — data sheet, stopwatch, counts. Note scheduling (MRP, hot lists, go-see).
- Map the current state — form, Assistant or CSV import; add information flows; check the mapping checklist on the Data tab is all ✓.
- Find the problems — load chart, issue tooltips, simulation findings; mark them with kaizen bursts.
- Design the future state — answer the eight questions; simulate it; it must ship on time on every simulated day.
- Set goals and limits — compare current, future and plan.
- Plan it — loops, owners, dates; suggest KPIs and log the baseline.
- Report — A3 print of current and future state, plus the plan; present in 10 minutes.
Facilitator guide: two-day workshop
For 6–12 participants in teams of 3, one laptop per team, the practice maps preloaded, and a real value stream that teams can walk on day 1.
| Time | Day 1 — see the current state |
|---|---|
| 08:30 | Welcome, goals, module 1 (foundations) with the takt exercise |
| 09:30 | Module 2: tour the app on the Brackets sample |
| 10:15 | Module 3: build the 4-step exercise map with the form |
| 11:00 | Gemba walk: collect data on the real stream (stopwatches, counts) |
| 13:00 | Teams map the real current state; facilitator checks the checklist |
| 14:30 | Module 4: reading the numbers; teams present their constraint and waiting days |
| 15:30 | Module 5: simulate; test two improvements |
| 16:30 | Debrief: what surprised you? |
| Time | Day 2 — design and plan the future |
| 08:30 | Module 6: the eight questions; teams design the future state |
| 10:30 | Module 7: goals and limits; plan within the limits; simulate |
| 12:00 | Module 9 (pick what fits: detail maps, mix, balance) |
| 13:30 | Module 8: implementation plan, owners, dates; KPIs with baselines |
| 15:00 | Teams print the A3 and rehearse |
| 15:30 | Presentations to leadership (10 min each) and commitments |
Tips: insist on floor data, not system data; keep teams on one family; time-box the future-state design (it gets better in the second pass, not the first); make the plan's first loop start within two weeks.
Quick reference
| Do this | How |
|---|---|
| Add a symbol | Drag from the palette, click it, or right-click empty canvas |
| Connect | Select, drag the green handle onto the next symbol |
| Style or reshape | Right-click a symbol or arrow |
| Drill into a process | Right-click → Create detail map; ⤵ opens it, ↑ goes back |
| Find problems | Load chart, hover % load / triangles, ▶ Simulate, 💡 Suggestions |
| Build the future state | Assistant → Design the desired state, or Copy as future state |
| Plan to a target | Goals → limits → Suggest changes → Build future state |
| Track | Plan tab (actions) and KPIs tab (measures) |
| Share / report | Save & share → link, file, synced file, versions, A3 / PDF |
Del delete · Ctrl+Z undo · Ctrl+Y redo · Ctrl+D duplicate · arrows nudge · F fit · Esc stop connecting
Formulas: takt = available time ÷ demand · days of stock = quantity ÷ daily demand · PCE = VA ÷ lead time · FPY = 100 − scrap − rework · load = need ÷ capacity · EPEI = Σ changeovers ÷ time left · pitch = takt × pack quantity · operators = work content ÷ takt.