Value stream mapping shows how material and information really flow from order to delivery, and where the lead time goes. How to map the current state, read it, design a future state and turn it into a plan that delivers.
The managing director was certain that the factory's biggest problem was machine capacity. A value stream map, drawn over three days with the team, showed something else. The product spent around 23 days from raw material receipt to shipment. The total processing time, when anything was actually being done to it, was under 40 minutes. The machines were not the constraint. The waiting was: in queues between departments, in weekly batching schedules and in a planning process that pushed work onto the floor whether the next step was ready or not.
That is what value stream mapping does. It shows the whole process, not just the parts each manager owns.
What a value stream is
A value stream is every step, value-adding or not, needed to take a product from order to delivery. It includes the flow of material and the flow of information that tells each step what to make and when. Most improvement efforts optimise individual steps. Value stream mapping looks at how the steps connect, because that is usually where the time goes.
Drawing the current state
Choose one product family: products that share most process steps. Then walk the process, starting at the shipping end and working back upstream. At each step record:
- Cycle time, changeover time, uptime and number of operators.
- Batch sizes and shifts worked.
- Inventory waiting before the step, in units and in days of demand.
- How the step knows what to make: a schedule, a kanban or an expedite.
Add the information flow from the customer, through planning and purchasing, to each process. At the bottom, draw the timeline: inventory days between steps and processing time at each step. Add them up to compare total lead time with value-adding time.
Reading the map
Most current-state maps reveal the same patterns:
- Value-adding time is a tiny fraction of lead time.
- Inventory piles up in front of steps with long changeovers or unreliable output.
- Several different schedules push work into the process independently.
- Information flow is complex, manual and slow, with expediting as the real control system.
The map makes the eight wastes visible across the whole flow, not just at one machine.
Designing the future state
A future-state map answers a set of standard questions:
- What is the takt time, the rate of customer demand?
- Where can we create continuous flow, moving parts one at a time between steps?
- Where must we use supermarket pull systems instead, because steps cannot be linked directly?
- Which single point in the process will receive the schedule (the pacemaker)?
- How will we level the mix and volume of production at the pacemaker?
- What improvements, such as changeover reduction, uptime or quality, are needed to make the future state work?
Those improvements are marked on the map as kaizen bursts, and they become the improvement plan.
Turning the map into results
A map without an implementation plan is just a poster. Break the future state into loops, prioritise them, assign owners and target dates, and review progress monthly. Typical first projects include kaizen events on changeovers, cells and replenishment loops.
Practical tips
- Map with the people who run the process, not from the office.
- Use pencil and paper first. Software comes later, if at all.
- Map a real product family over a real period, not an idealised process.
- Repeat the exercise once the future state is achieved. It becomes the next current state.
Value stream mapping is not complicated. Its power is that, for the first time, everyone sees the same picture of where time and money actually go.
