Value Stream Mapping: From Factory Map to Better Flow

Home » Resources » Value Stream Mapping: From Factory Map to Better Flow

Estimated reading time: 12 minutes

A part can spend minutes being worked on and days waiting to move. Drawing the operations in order will show where it goes. Value stream mapping helps you understand why it takes so long to get there.

Value stream mapping (VSM) is a lean method for showing the material and information flows needed to deliver a product, including the work, inventory and delays along the way. You map the current state, design a better future state and agree on the changes needed to reach it.

For a manufacturing team, the useful result is a decision: which change will improve the flow to the customer? Sometimes that means a physical change, such as reducing a setup. Sometimes it means stopping the release of work that will only join a queue. Where the problem is competing orders on shared resources, SkyPlanner APS turns the agreed operations and priorities into a coordinated production schedule, reducing the need to rebuild those queues by hand.

What a value stream map shows that a process flowchart misses

A process flowchart can describe the sequence: cut, form, assemble, dispatch. A value stream map connects that sequence to the conditions controlling it. How much work is waiting? How often does planning send instructions? Does a supervisor follow the schedule or a separate urgent-job list?

The map brings three views together:

• Material flow: the processes, stock locations and movements from the chosen starting point to the customer-facing endpoint.
• Information flow: the orders, schedules, replenishment signals and other instructions that tell each process what to do.
• Time and operating data: process times, changeovers, available time, quality, inventory and delays that explain the flow.

A routing describes how to manufacture the product. A value stream map asks how the whole stream currently performs and how it should work instead. Include activities that do not add customer value; leaving them out would hide the very improvement opportunities you want to see.

1. Choose one product family and a clear boundary

Start with products that follow similar processing steps and use similar resources. A simple product-by-process matrix can reveal which products belong together. Do not put every factory product on the first map just because they all pass through dispatch.

Choose a family with a worthwhile problem: long delivery lead times, excessive work in progress, recurring shortages at assembly or repeated expediting. Bring together people who know production, planning, materials and quality. Their different views help explain why work stops between departments.

Write the scope on the map. For example: “Repeat enclosure family, from incoming blanks to finished goods ready for dispatch.” Record the customer demand, product mix and observation period. If order entry or an external treatment causes important delays outside that boundary, show the interface and investigate it rather than silently treating it as zero time.

A mixed-product job shop can still use VSM. Select a representative family or route pattern, record meaningful variation and identify the machines shared with other families. An average that combines a two-minute job with an eight-hour job may describe neither well enough to guide a change.

2. Walk the current flow and collect actual data

Follow the material on the shop floor with the people who do the work. Walking upstream from dispatch can help you see what each process needs from the one before it. Draw the actual sequence, including subcontracting, rework loops and stock locations where relevant. Use the ERP record as supporting data, not as proof that the floor operates exactly as documented.

For each process and the gaps around it, capture enough information to explain performance:

What to record	Why it matters
Measured cycle or processing time, with its unit and quantity basis	Separates the work itself from time spent waiting.
Setup time and actual production/transfer batch sizes	Shows whether changeovers or waiting to complete a batch influence flow.
Available working time, shifts, people and resource sharing	Helps explain why an apparently quick operation may still have a long queue.
Inventory quantities and their status	Distinguishes usable stock from incomplete, rejected or quarantined material.
Quality losses and rework	Makes repeated work and interrupted flow visible.
Arrival, start and completion times for representative orders	Checks how long work actually waits and travels through the chosen boundary.
Source, frequency and recipient of production instructions	Reveals how work is released and how priorities change.

Label the date, source and basis of the data. A single morning’s stock count is a snapshot; a representative average from repeated observations is stronger evidence for a recurring condition. Where products differ substantially, retain a range or split the family rather than hiding the variation in one number.

Ask an operator what determines the next job. If the answer is a phone call, mark that instruction even if the official process says “daily ERP schedule.” This is often where the map begins explaining a delay instead of merely displaying it.

3. Draw the material flow, instructions and timeline together

Place the customer and supplier at the edges, with the main processes in sequence between them. Add the inventory locations, quantities and movement links. Above the process flow, show how customer requirements reach planning and how instructions reach production. Put the relevant process data beneath each process and the time ladder along the bottom.

Use familiar VSM symbols, with a small legend so everyone reads them consistently. A process box represents a process; an inventory triangle marks stored material or WIP. Material arrows show movement. Information arrows show instructions, with different line styles where manual and electronic communication need distinguishing. A data box carries operating figures, and a kaizen burst marks a proposed improvement.

In a future-state map, a supermarket symbol denotes controlled replenishment stock, while a FIFO lane denotes a defined first-in, first-out connection. Those symbols should represent agreed operating rules, not decorative ways to connect boxes.

A compact manufacturing example

Imagine a repeat enclosure family with average customer demand of 100 good enclosures per production day. The current-state material flow is:

Incoming blanks → Cut → Form → Assemble → Finished goods ready for dispatch

The map also records the waiting stock and measured processing times:

Map location	Waiting inventory	Processing time per enclosure
Before cutting	100 units	Cutting: 20 seconds
Between cutting and forming	300 units	Forming: 40 seconds
Between forming and assembly	100 units	Assembly: 60 seconds
Finished goods awaiting dispatch	100 units	Outside this example’s processing-time total

Above that material flow, draw the actual information flow: customer requirements reach planning daily, planning sends separate lists to cutting and forming, and assembly requests urgent parts by phone. Cutting starts large batches whenever it has spare time. Forming shares a machine with other product families.

Now the map has something to explain. The largest queue is not necessarily at the slowest measured operation. It may be where early releases meet competition for shared capacity.

4. Interpret the waiting without confusing the time measures

For a stable, comparable product flow, a useful first estimate of inventory coverage is:

Inventory coverage in production days = inventory quantity ÷ average daily demand

In the example, the queue before forming represents 300 ÷ 100 = 3 production days of demand. The four waiting-stock locations together hold 600 units, equivalent to 6 production days at that demand rate.

The three processing times add up to 20 + 40 + 60 = 120 seconds, or 2 minutes per enclosure. That contrast tells the team to investigate waiting before concentrating only on faster machine cycles.

But six days of inventory coverage is not a stopwatch measurement of every order’s lead time. Product mix, changing demand, batch movement, operating calendars and priority overtaking can all affect the actual elapsed time. Validate the picture against representative order timestamps, keeping the starting and ending points consistent. Do not quietly substitute calendar days for production days.

Nor are two minutes of processing automatically two minutes of value-added work. In lean terms, value-creating work changes the product in a way the customer values. Inspection or rework may occupy time without meeting that definition. If you calculate a value-added ratio, classify the activities first and use the same time basis for numerator and denominator.

Compare the process with demand, not just with its neighbour

If the family has 420 available production minutes per day, its demand-based takt time is 420 ÷ 100 = 4.2 minutes per enclosure. Takt expresses the required customer pace, not the measured speed of a machine.

The example’s individual processing times are below that pace, yet the map still shows a large queue. That does not prove the resources have spare capacity: the forming machine also serves other families, and setups, availability and staffing consume time. Check the whole load before declaring a bottleneck or buying another machine.

Making cutting four seconds faster will not, by itself, explain or remove three days of stock before forming. The more useful question is why those 300 units were released and what prevents them from moving onward.

5. Design a future state around the cause of the delay

Use the current map to agree how material and information should flow, not simply to draw smaller inventory numbers.

For the enclosure family, investigate the forming queue with the planner and operators. Are usable parts waiting for capacity? Are jobs blocked by missing material or a quality release? Does the daily list keep changing? Does cutting produce ahead because its performance is judged by local machine utilization?

Different causes need different changes. Reducing setup time can make smaller batches practical. Connecting suitable processes in continuous flow can remove an intermediate queue. Where that is not feasible, a controlled FIFO connection or supermarket replenishment may regulate what moves next. Necessary curing, cooling or quality holds must remain visible until an approved process change removes the need for them.

For a repeat-flow design, decide where the customer schedule enters the stream. This is often a pacemaker process: the point whose production rhythm governs the stream. Define how upstream work is authorized and how downstream work follows it. Sending independent, conflicting schedules to every process can recreate the queues the future map was meant to remove.

Suppose the team sets a pilot target of reducing the waiting stock before forming from 300 to 100 units. That would remove two production days of inventory coverage at the same demand rate. It is a target to test, not an automatic two-day delivery improvement.

Make the operating change explicit: what authorizes cutting, how much may wait before forming, who handles an urgent exception and how forming’s capacity is coordinated with other products. Smaller triangles on paper cannot enforce those decisions.

Put scheduling improvements into practice with SkyPlanner

The future-state map defines the intended flow. SkyPlanner helps make the scheduling part work across the real order book, rather than only within the family you mapped.

Represent the agreed operations as product process steps, with eligible workstations, processing durations and setup times. Record shifts and resource availability, then set the priorities that reflect your delivery commitments. Arcturus, SkyPlanner’s scheduling AI, calculates the schedule against capacity, operation dependencies and those priorities. With material-aware scheduling enabled, material availability also affects when work can run.

For the enclosure family, that means forming work is scheduled alongside the other orders competing for its machine. You can inspect the resulting sequence in the Gantt view instead of maintaining a separate family plan that assumes the machine is always available. The benefit is less manual queue reconciliation and earlier visibility of conflicts that would otherwise leave parts waiting.

Keep the flow-control decision intact. If your future state uses pull replenishment, the signal authorizes the work; scheduling fits the authorized work to available resources. Do not turn a capacity calculation into permission to manufacture unwanted stock.

When an urgent order or a shift change alters the situation, Arcturus can recalculate the schedule. Your rules and locks protect decisions that should remain stable. SkyPlanner integrates with any ERP or MES through its open REST API, so you can improve scheduling while retaining the systems that already hold your orders and production data.

What coordinated scheduling changed for Mamec

Mamec, a Finnish manufacturer of mixing technology and high-tech contract products, had thousands of interdependent work orders. Supervisors prioritized jobs from delivery dates on paper orders, while management lacked a reliable overall view of load.

The company integrated SkyPlanner with Oscar ERP. AI-powered finite scheduling now handles the scheduling groundwork, and work queues locked a week at a time help machinists know and prepare the next jobs.

In its SkyPlanner experience, Mamec reports less workstation downtime from a better work sequence, a lighter planning workload and less working capital unnecessarily tied up in WIP. Its experience illustrates the practical value of coordinating interdependent work: an improved flow needs a shared plan people can keep using as production changes.

Start with one change you can verify

Give the first future-state action an owner, a completion date and a measurable outcome. For the forming queue, track waiting inventory and actual arrival-to-start time alongside throughput, delivery performance and urgent exceptions. Check that a smaller queue has not merely moved the delay to cutting, assembly or the customer.

Review the result with the people who operate the stream. Keep what improves flow, investigate what did not work and update the current-state map before choosing the next change.

To see how the scheduling part would work in your factory, book a SkyPlanner demo. Bring the mapped family, its largest queue and the other jobs sharing that resource. That is a concrete starting point for turning a useful map into less waiting and a production plan your team can follow. If you prefer to explore first, try SkyPlanner with demo data.