Pillar 02 · The Hidden Economics of Manufacturing

All KPIs derived in the end-of-day reports are lagging indicators. Most of the shop floor looks normal at a high level in those reports. But when you walk through the floor, things look very different.

The report may show that the machines are running, production is progressing and the factory is close to its planned capacity. Everything may appear reasonably normal. But when you walk through the shop floor, you begin to notice things that do not appear with the same clarity in the reports.

There are racks containing parts that are neither finished goods nor active production. There are trolleys waiting for inspection. There are components separated because of a small dimensional variation. There are batches waiting for a decision on whether they can be reworked. Some have been there for a few hours. Others have been there for several days.

They are not necessarily scrap. They may not even be classified as a problem. But they are no longer moving through the factory in the way they were supposed to.

I have seen this repeatedly in manufacturing. A production plan may look reasonably healthy while the shop floor tells a different story. There can be a significant amount of unfinished material sitting between processes, waiting for inspection, waiting for rework, waiting for a decision or simply waiting because nobody is quite sure what should happen next.

I call this hidden WIP.

The problem with hidden WIP is not simply that inventory is sitting on the floor. The bigger problem is that the factory has already spent resources on those parts, while the value expected from those resources has not yet been realised.

A component sitting in a rack has already consumed raw material. It may have consumed machine time, operator time, electricity, tooling, inspection time and floor space. If there is a quality issue, more resources will be required before the part can become a good part. The customer, however, does not receive two parts because one required rework. The customer still expects one good part. That is where the economics starts becoming interesting. Research on rework costing in manufacturing has similarly shown why rework needs to be treated as a measurable production cost rather than only as a quality statistic.

The cost of making the same part twice

Consider a component that goes through a machining or molding process and is later found to have a minor quality issue.

At first sight, the issue may look small. Perhaps a dimension is outside the required tolerance. Perhaps there is a cosmetic defect. Perhaps a process parameter was not within the expected range. Perhaps the part simply needs another operation before it can be accepted.

The immediate discussion normally starts with the material. How much material was lost?

That is a reasonable question, but it is often not the most important one. The part has already occupied a machine. An operator has already handled it. Energy has already been consumed. Tooling has already been used. Inspection has already taken place. Production planning has already allocated time for it. Now the same part needs another round of attention. The factory therefore uses additional resources to produce the same saleable output.

This is where rework becomes much more than a quality problem. It becomes a capacity problem.

The economics becomes even less obvious in processes where some of the material can be recovered.

Take injection molding as an example. Depending on the material, product and process requirements, runners, sprues and even rejected components may be ground and returned to the production process. From the perspective of raw material, some of the apparent loss can therefore be recovered.

But the machine cycle cannot be recovered. The operator time cannot be recovered. The electricity consumed during the cycle cannot be recovered. The mould occupancy cannot be recovered. The inspection and handling effort cannot be recovered. Most importantly, if that machine is already a constraint, the production opportunity that was lost during that cycle cannot be recovered.

The material may come back. The capacity does not.

This distinction is easy to miss because material is visible and capacity is not. A rejected component can be weighed. Its material cost can be calculated. The rework cost can be estimated. But the economic value of the machine hour that was consumed is much harder to see.

The factory within the factory

Once rework begins to accumulate, another phenomenon appears. The factory develops a second flow.

There is the production flow that everybody understands:

Production → Inspection → Finished Goods → Shipment

Then there is another flow that develops around it:

Production → Quality Issue → Waiting → Decision → Rework → Inspection → Waiting → Production

This second flow is rarely designed as part of the manufacturing system. It emerges because something went wrong in the first flow. Yet it requires people, machines, space and management attention. This is what makes the idea of a hidden factory so important. Manufacturing research has used the term to describe the rework, process inefficiencies and additional overhead created by problems within the production system. NIST has specifically noted that such hidden-factory effects can have significant long-term impacts on manufacturing operating costs. The American Society for Quality also treats rework, scrap and other internal and external failures as part of the economics of quality.

The hidden factory does not necessarily have another building.

• It can be a few racks beside a machine.
• It can be an operator spending part of the shift correcting an earlier problem.
• It can be an inspection queue.
• It can be a planning team repeatedly changing schedules because yesterday's production did not come through as expected.
• It can be a machine producing components that will later have to pass through another process before they can be shipped.

And because this work is distributed across the factory, it is easy for nobody to see its total economic impact.

When yesterday's problem consumes tomorrow's capacity

There is another reason hidden WIP is dangerous. It does not stay in yesterday. The parts that are waiting today eventually compete with tomorrow's production.

The same machine may be required to complete the old rework and the new customer order. The same operator may be required to handle both. The same inspection resource may have to clear both queues. What looked like a small quality problem yesterday can therefore become a scheduling problem tomorrow.

This is where many factories begin to feel that they have a capacity problem even though the machines appear to be running.

The problem is not always a shortage of machines. Sometimes the capacity already exists, but a portion of it is being consumed by work that should not have required that capacity in the first place. NIST has documented manufacturing examples where reducing WIP and improving flow resulted in substantial increases in throughput and available capacity. In one recent case at Argon Medical, overall process WIP was reduced by 75% while process capacity increased by 125%, with throughput more than doubling in two value streams (NIST MEP success story).

The important point is not the particular lean technique used there. It is that capacity can be hidden inside the way work flows through the factory.

Not every hour of capacity has the same value

This becomes even more important when we look at the constraint. Suppose a factory has ten machines and one of them determines the maximum output of the production line. Losing an hour on one of the other machines may create some inconvenience. Losing an hour on the constraint can affect everything downstream. The economic value of capacity is therefore not uniform across the factory.

An hour of unused capacity on a non-constrained resource may not have much immediate economic consequence. An hour consumed unnecessarily on the machine that limits production can be very different. This is why looking only at machine utilisation can be misleading. A machine may show 90 percent utilisation and still be producing less value than expected because part of its time is being consumed by rework, repeated processing, waiting between activities or slower-than-expected cycles.

The machine is busy. But being busy is not the same as creating value.

A machine that should complete a cycle in 45 seconds but consistently takes 48 seconds may never appear as downtime. It may remain operational throughout the shift. Yet across thousands of cycles, those three seconds accumulate into a significant amount of lost production time. The same is true of small interruptions, repeated setups, quality checks, material waiting and unnecessary movement. Individually, they are easy to ignore. Together, they can change the economics of the factory. NIST's work on manufacturing inventory and flow time similarly treats rework and defects, inventory and waiting as forms of waste that can be understood by looking at how material and time move through the manufacturing system (NIST publication).

The hidden cost is often not a cost

This is the part that I find most interesting.

When management talks about manufacturing losses, the conversation usually goes towards cost.

Material cost, labour cost, energy cost, maintenance cost, scrap cost and rework cost.

These are important, but there is another question that is often missed.

What revenue could this capacity have produced?

Imagine a factory where a particular machine is already running close to its practical limit. A new customer approaches with a significant order. The factory looks at its available capacity and concludes that another machine, another shift or another investment may be required.

That decision may be completely justified. But before investing in new capacity, it is worth asking another question.

• How much of the existing capacity is actually producing saleable output?
• How much is being consumed by rework?
• How much is sitting in WIP?
• How much time is lost between processes?
• How much capacity is being consumed by slow cycles?
• How much production time is spent correcting problems that originated earlier in the process?

The answer can sometimes change the investment conversation completely. This is not an argument against investing in new machines. There are many situations where new capacity is absolutely necessary. The point is that the economic value of a new machine cannot be understood independently of the capacity already available in the factory. If existing capacity is being lost inside the process, adding more capacity may simply increase the size of the system without addressing the underlying loss.

There are several documented examples of manufacturers creating additional sales by improving the capacity of their existing operations. At Nevada Automotive Test Systems, for example, process improvements increased production capacity and reduced lead times, enabling the company to increase annual sales by 58 percent without increasing headcount (NIST MEP success story).

In another NIST-documented example, improving production flow at Wallbeds by Wilding reduced the time required for a unique product request from one week to one day and increased production capacity, contributing to new and retained sales (NIST MEP success story).

These examples point towards an important distinction. Operational improvement is not only about reducing today's cost. It can also create the capacity to generate tomorrow's revenue.

This is why hidden economics matters

The factory rarely loses its economics through one dramatic event. It loses it gradually.

• A few parts waiting for inspection.
• A batch waiting for a quality decision.
• A rework operation added to the process.
• A machine running slightly slower than its designed cycle.
• An operator correcting yesterday's problem.
• A bottleneck spending its available hours on work that will have to be processed again.
• A trolley of WIP occupying space and attention because nobody is certain how old it is or what should happen to it.

Each one may be too small to become a management issue on its own. But manufacturing is a system. Small losses interact with one another.

A quality problem creates rework. Rework creates additional WIP. WIP creates waiting. Waiting affects planning. Planning changes the schedule. Schedule changes create additional setups. The constraint loses productive time. Delivery pressure increases. Overtime follows.

By the time the financial impact becomes visible, the original problem may be several steps behind it.

This is why the economics of manufacturing cannot be understood only through the traditional cost structure. There is another economy inside the factory. An economy of time, capacity, flow and opportunity. And much of it remains hidden until somebody follows the material through the complete process.

Making the invisible visible

Digital transformation can play an important role here, but not simply by putting more dashboards on the wall. The objective should be to connect the pieces that are normally separated.

A quality issue should not remain only a quality record. It should be possible to understand which machine produced it, which material was used, which process conditions existed, how much time was consumed, whether rework is required, where the part is now, how long it has been waiting and which customer order depends on it.

Similarly, WIP should not be treated only as a quantity. Its age, status, reason for waiting and relationship to production capacity can be much more important than the number of pieces itself.

The real value of data comes when these relationships become visible.

A factory should be able to move from:

“We have 500 pieces of WIP.”

to:

“These 500 pieces represent 140 machine hours, 36 of which are waiting because of quality decisions, and 72 are competing with next week's customer orders for the same constrained resource.”

That is a very different level of understanding. It changes the conversation from inventory management to economics.

Looking at a loss differently

This is how I use the Hidden Economics of Manufacturing Canvas. One real loss, followed through the system, from what was already consumed to the capacity it is still holding.

The Hidden Economics of Manufacturing Canvas

Download the canvas (PDF) →

The canvas helps you follow one real operational loss through the system:

• Start with the loss itself. What is actually happening?
• Then ask what has already been consumed—material, labour, machine time, energy, inspection, space and management attention.
• Then look at what capacity is being consumed. What process is affected? Is it a constraint? Is the capacity needed for something else?
• Then follow the consequence downstream. Has the loss created WIP? Has it created rework? Has it delayed another order? Has it changed the production plan? Has it affected delivery?
• Only after following the loss through the operation should we ask about the economic consequence. What did the loss actually cost? More importantly, what value could the recovered capacity create?

This changes the way a manufacturing problem is viewed.

Instead of asking only:

“How much did this defect cost?”

we start asking:

“What capacity did this defect consume, and what could that capacity have produced?”

That is where the hidden economics starts becoming visible.

The economics of tomorrow

The losses that worry me rarely show up as their own line on the statement. A trolley of WIP. A rework loop nobody scheduled. A cycle a few seconds slow, all shift, on the machine that limits the line. Yesterday's profit can still look acceptable, so the factory lives with them. The bill arrives later, as an order you could not take, a shift you had to add, or a machine you bought earlier than you needed to.

Recovering material, scrap and labour hours still matters. When that recovered time is on the constraint, the factory also gets room to grow before it buys the next machine. Many of these costs do not reduce today's profit. They reduce tomorrow's ability to take the next order.

How much capacity is already in your factory?

Next time you walk the line, pick one trolley that has been sitting there since yesterday. Write down what that material already consumed — machine time, labour, energy, a quality decision still open — and which machine it will steal time from tomorrow. That note is more useful than the scrap percentage in the month-end pack.

The Hidden Economics of Manufacturing Canvas is a place to write it down, if you want the questions on one sheet.

How much of tomorrow's capacity is already spoken for by work that should have been finished yesterday? That is where I would start.