Value Stream Mapping for a Small Factory: See Where the Time Goes

What is a value stream map?

A value stream map is a one-page picture of the whole journey of one product family, from the material arriving at your gate to the finished goods leaving on the truck. It shows where the material goes, how orders reach each machine, and how long the part waits compared with how long it is worked on.

Many owners say the same thing. Every machine is running, and orders are still late. There are crates everywhere, but the part we need is not there. The supervisor spends the day chasing jobs.

When we look for the cause, we look inside one machine, a slow lathe or a weak operator. But in most plants the problem is not in the machines. It is in the journey between them: the waiting, the piles of stock, and the way orders reach the floor.

How is it different from a flowchart?

A flowchart shows the steps. A map shows the time.

Who should make the map?

A small team of five to seven people, on the floor:

• a sponsor, usually the owner or plant head, who gives time and decides on changes;
• one person who owns the whole journey, from order to dispatch, not one machine;
• planning, the supervisor, one operator from each step, a person from stores and one from quality.

You need a pencil, an eraser, a stopwatch and last month's records of downtime, production and rejection. No computer. Drawing by hand makes everyone look at the floor, not at a screen.

How do you draw it, step by step?

1. Choose one product family. Products that go through the same machines. Start with the one that hurts most: the biggest volume, the most rejection or the most stock. Do not map the whole factory on one sheet.
2. Walk the floor, from dispatch back to the store. Walking backward keeps the customer in mind. If the office data and the floor do not match, the floor is right.
3. Draw a box for each process, with a data box under it: cycle time, changeover time, uptime, batch size and operators. Time the cycle yourself with a stopwatch. Take the usual time, not the best.
4. Count the stock between the boxes with your own eyes, and change it to days.
5. Draw how orders flow. Who tells each machine what to make? A weekly list, a phone call, a paper slip?
6. Draw the timeline: days of waiting on top, minutes of work below.

Days of stock = pieces waiting ÷ pieces the customer takes per day.

Example: where does the time go?

These figures are an example for a machine shop making one shaft family. The customer takes 25 pieces a day.

That is about 25 days of waiting. The actual work on one shaft adds up to only about 21 minutes. In a 450-minute shift, over 25 days, the work is less than 1% of the time. This is normal in a plant that has never mapped. It is not a reason for shame. It is a reason for hope, because the waiting is so large that simple changes make a big difference.

Also compare each machine's cycle time with the customer's pace. If the customer needs one piece every 18 minutes and no machine takes more than 8, the machines are not the problem. The waiting is. Many owners want to buy a faster machine when the real problem is stock, batches and scheduling.

What does the map usually show?

Marked as bursts on the map, the problems in an example like this are:

• a two-hour changeover on one machine, so the plant runs big batches;
• every machine gets its own list, so each pushes its batch whether the next step is ready or not;
• too many days of finished stock, in the wrong sizes;
• two big material deliveries a month;
• a supervisor who chases all day because priorities change daily.

None of these is "the operator is slow." The map points at the system.

The map is not the goal. The change on the floor is.

How do you draw the future map?

The future map shows how the plant should run in three to six months, with the same machines and people and less waiting. Ask six questions:

1. Do we ship from a small finished store, or direct to the truck?
2. Where can parts flow from one machine to the next without stopping?
3. Where do we need a small store, refilled by pull?
4. Which one point gets the signal to start? Only one point should be scheduled. The rest follow.
5. What must improve to make it work: changeover, breakdowns, quality?
6. What is the plan, with an owner and a date for each change?

Order matters. Cut the changeover time first. If you cut batch size before changeover, the machine spends the day in setups and output falls.

How to start in your plant

1. Choose one product family.
2. Walk it from dispatch to the store, with a stopwatch.
3. Count the stock between the steps and change it to days.
4. Mark the five biggest problems on the map.
5. Give each one an owner, a date and a measure.
6. Review weekly for 30 minutes on the floor, and redraw the map after six months.