Pillar 01 · The Responsive Factory

In March 2000, a fire broke out at a Philips semiconductor facility in Albuquerque, New Mexico. The fire itself was not what made the incident remarkable. What followed was. The facility supplied components to two of the world's leading mobile phone manufacturers, Nokia and Ericsson. What initially appeared to be a temporary disruption became a prolonged supply problem because contamination prevented the facility from returning to normal production as quickly as expected. Ericsson later described the incident in its own corporate reporting as a severe shortage of integrated components for mobile handsets, caused in part by the absence of an alternative supplier.

Both Nokia and Ericsson were exposed to the same event. Both depended on the same supplier. Both were operating in the same rapidly expanding mobile phone market. Yet the consequences for the two companies were different.

Nokia moved quickly to understand the seriousness of the disruption. It worked with Philips, explored alternative sources and redesigned components where necessary. Ericsson also attempted to respond but slowly, but the prolonged component shortage affected its production significantly. Ericsson's subsequent account of the period describes the disruption as a severe component shortage that adversely affected its operations. The lesson is not that one company had a perfect supply chain and the other did not. The more interesting lesson is that when an unexpected event occurs, the ability of an organization to sense the change, understand its implications and act quickly can determine how much of the disruption ultimately reaches the customer.

That distinction is becoming increasingly important in manufacturing.

For a long time, the dominant manufacturing question has been how efficiently an organization can execute a plan. Factories have invested enormous effort in reducing waste, increasing utilization, improving OEE, reducing changeover time, standardizing work and increasing throughput. These disciplines remain essential. But they assume something that is often no longer true: that the plan will remain reasonably stable.

Markets do not always behave that way.

Customers change priorities. Demand moves unexpectedly. Product mixes change. Suppliers miss commitments. Machines fail. New products have to be introduced into existing production systems. Quality problems interrupt production. Energy availability and cost can change. A disruption somewhere in the supply chain can suddenly alter what a factory is capable of producing.

At that point, the question changes.

It is no longer only about how efficiently the factory can execute the plan. It becomes a question of how quickly the factory can change when the plan itself changes. That is the idea behind the Responsive Factory.

The difference between efficiency and responsiveness

Efficiency and responsiveness are sometimes treated as if they are the same thing. They are not.

A factory can be extremely efficient at producing a stable product mix in a stable environment and still struggle when demand changes. It can have excellent machine utilization and yet take two days to reorganize production after an important customer changes its requirement. It can have sophisticated planning software and still depend on several people manually interpreting information before a decision can be made.

Efficiency asks whether the operation is performing well against an established expectation. Responsiveness asks what happens when that expectation is no longer valid.

This is particularly important because flexibility itself has economic value. McKinsey has used Honda's East Liberty, Ohio plant as an example of manufacturing flexibility, noting that the plant could switch in minutes between producing the IC engines and the EV vehicles. The significance is not simply that two models can be produced in the same facility. It is that the production system was designed to give the business more options when demand or product requirements change.

That changes how we think about capacity.

A traditional capacity discussion might ask how many units a factory can produce in a month. A responsiveness discussion asks a slightly different question: how much of that capacity can actually be redirected when the business needs something different?

There may be capacity available on paper, but if it cannot be mobilized quickly because of tooling, skills, material, sequencing, planning or organizational constraints, that capacity is not necessarily useful when the moment arrives.

This is one of the hidden dimensions of manufacturing performance.

Faster can make the factory worse

It is tempting to read responsiveness as “doing things faster.” A factory that changes its schedule every time a new signal appears may simply be unstable. Constantly changing priorities can create more setup time, more WIP, more confusion and, in the end, worse delivery. The organization has to recognize which changes matter, understand what they do to the plan, and then decide what actually needs to move.

Consider a sudden increase in demand from an important customer. The first signal may arrive through sales. But that information by itself is not enough. Someone needs to understand whether the additional demand can actually be fulfilled. Planning needs to look at existing orders and available capacity. Procurement needs to understand material constraints. Production needs to know whether the required machines, people and tooling are available. Management may need to decide which existing commitments should be protected and which can move.

The response is a chain: sense, understand, decide, execute, and then recover. How fast that chain moves depends on the machines, and also on how fast information and decisions move through the organization. A responsive factory is one where those stay close enough that the operation can change and still keep control.

Responsiveness can be designed into the operating model

The automotive industry provides some of the clearest examples of this thinking.

Mercedes-Benz has described its Factory 56 operation in Sindelfingen around flexibility in production models, volumes and material flow. The company designed the production system so that different vehicle configurations and drive technologies could be integrated into the manufacturing environment and production could be adjusted according to market requirements.

BMW has similarly described flexibility across its production network, including the ability to manufacture different drivetrain types and model variants on production lines and adjust volumes according to market conditions.

These examples are interesting because they demonstrate that responsiveness does not have to be improvised when disruption occurs. It can be designed into the factory before the disruption arrives.

The factory layout can provide flexibility. Equipment can be selected for flexibility. Product architecture can make production easier to change. People can be cross-trained. Planning systems can provide alternative scenarios. Material strategies can provide options. Decision rights can be clarified.

What you get is a factory with more choices when it needs to change, and the judgment to leave the rest of the schedule alone. Reacting to everything is just noise.

Sometimes the constraint is not on the shop floor

This is where responsiveness becomes particularly interesting for senior management.

Imagine that a machine fails at 8:00 in the morning.

The operator knows immediately. The maintenance team knows shortly afterwards. Production understands that there is a problem. But perhaps the planning team does not know the full consequence until later in the morning. Perhaps the plant manager is informed only when the missed production begins to threaten a customer commitment. Perhaps the decision to move the work to another machine requires an approval.

The machine may have been unavailable for two hours. But the business may have taken six hours to respond to the consequences of that failure.

The physical downtime and the organizational response time are two different clocks. A machine breakdown is easy to see. The time spent waiting for information, for a decision, for a rebuilt schedule, for material, for an approval, is much easier to miss, and that is often where a large part of the loss sits.

Responding to constraints beyond the factory

The global semiconductor shortage of 2021 provided another illustration.

For automotive manufacturers, the problem was not simply that there were fewer chips available. Once supply became constrained, manufacturers had to make decisions about which vehicles to build, which plants should receive available components, which products should receive priority and how production should be adjusted.

General Motors, for example, announced production cuts at several plants and said it would prioritize production of higher-margin vehicles during the shortage.

This illustrates an important shift. When resources are abundant, responsiveness can appear to be an operational issue. When resources become constrained, it becomes a strategic allocation issue.

The question moves from “Can we produce?” to “What should we produce, where, and which commitment gets the parts?” That is a different kind of manufacturing decision.

The five dimensions of the Responsive Factory

From these situations, I see responsiveness as five connected capabilities.

1. Demand. The organization needs to recognize that something has changed. This might be a change in customer demand, a change in product mix, a supplier problem or an internal operational event. The important question is not simply whether the information exists somewhere in the organization, but how quickly the right people become aware of it.
   
   
2. Decision. Once the change is understood, somebody has to decide what should happen. This is often where organizations lose surprising amounts of time. Information may be available, but decision ownership may be unclear. Multiple approvals may be required. Different departments may interpret the situation differently. The factory cannot respond faster than the organization can make the necessary decisions.
   
   
3. Capacity. A decision has to be translated into something physically achievable. What machines are available? What WIP exists? Which materials are available? Which operations are constrained? What changeovers are required? Can another machine or plant take the work? Capacity therefore needs to be understood as usable capacity under current conditions, rather than simply a number on a planning report.
   
   
4. Execution. This is where the revised decision reaches the physical operation. People need to know what changed. Materials need to move. Machines need to be prepared. Quality may need to release the revised process. The revised priority needs to become reality on the shop floor.
   
   
5. Recovery. A responsive factory does not stop at the first reaction. It also needs to recover from the change. If an urgent order is inserted into the schedule, what happens to everything else? If a machine failure is recovered, how does the factory restore the production sequence? If a material shortage is resolved, how quickly does the operation return to a stable state?
   
   

The chain is Demand → Decision → Capacity → Execution → Recovery. Under it sits a plain requirement: information has to move faster than the problem.

Where does responsiveness break down?

I am less interested in whether a factory calls itself responsive, and more interested in where the time disappears. Good data with slow decisions. Flexible machines with a rigid plan. Enough capacity, and no view of it. Skilled people, and no way to move them. An ERP on the wall, and a spreadsheet plus a phone call when something important changes.

That gap can sit between sales and planning, planning and procurement, procurement and production, production and quality, or the plant and logistics. Looking at one department on its own will not show it. Responsiveness runs from the customer to the finished product.

Looking at your own factory

This is where I would introduce the Responsive Factory Canvas.

I would not ask an organization to begin with a generic maturity score or a questionnaire containing fifty questions. Those exercises can be useful, but they often tell management what it already suspects without showing where the time is actually being lost.

Instead, start with one real event.

Choose something that happened within the last three months. A significant customer priority change. A machine breakdown. A material shortage. An unexpected demand increase. A quality issue. A product change. Something that forced the organization to respond.

Then reconstruct what actually happened.

• When did the event occur?
• When did somebody first know about it?
• When did the organization understand its significance?
• When was the decision made?
• When did the production plan change?
• When did the shop floor actually begin responding?
• When did the operation recover?
• And what did the elapsed time cost the business?

This simple exercise can be surprisingly revealing.

Imagine that a customer changed a priority at 9:00 AM. Sales understood it at 9:15. Planning became aware at 11:00. A revised schedule was ready at 2:00 PM, but material was not available. The material arrived the next morning, and production finally changed at 4:00 PM.

The factory might initially describe this as a material problem.

But the canvas reveals something more useful.

The organization had a 31-hour response cycle.

Now the conversation can move beyond “Why didn't we have the material?”

Management can ask how much of those 31 hours came from information delay, how much came from decision-making, how much came from planning, how much came from material availability and how much came from execution.

That is a much more powerful conversation.

The Responsive Factory Canvas

Download the canvas (PDF) →

The canvas therefore follows the actual journey of the event:

• What changed?
• When did we know?
• When did we understand the impact?
• When did we decide what to do?
• When did the plan change?
• When did execution change?
• When did we recover?
• What did the delay cost?

The canvas is there to make the response visible, not to hand the factory a score. Map one incident, then another. After a few, a pattern usually shows up.

Perhaps decisions are always slow because ownership is unclear. Perhaps planning is consistently disconnected from real capacity. Perhaps material availability is repeatedly the constraint. Perhaps production has flexibility that the planning system cannot see. Perhaps the data exists but arrives too late.

At that point the plant is looking at how it is built to respond, and a conversation about process, data, planning or technology has something real to attach to.

From responsiveness to transformation

I don't separate digital transformation from the way the plant already runs. A factory can buy machine data, a planning system, an MES, analytics or AI and still be slow. The technology has to shorten the distance between an event and a useful decision. If a machine problem is detected immediately and the production team hears about it an hour later, the real-time signal did little. If a planning system can rebuild a schedule in seconds and the approval takes a day, the time was not in the scheduler. If the data is on a screen and nobody has agreed what to do when a number moves, another dashboard will not speed the response.

What I want from a digital project is less time between the moment the factory changes and the moment someone acts on it with a clear head.

How responsive is your factory?

Take the last time a customer changed a priority. Write the times: when it happened, when planning heard, when the schedule moved, when the line actually changed. The gap between those times is the problem. The software discussion can wait until you can see the gap.

The Responsive Factory Canvas is a sheet for those times, if you want them in one place.

The Responsive Factory

The plan will keep changing. Customers change requirements, machines fail, a supplier misses a commitment, and someone on the floor has to absorb it before the day is over. Efficiency still matters. It tells you how the factory performs when the assumptions in the plan hold. Responsiveness is what you have left when they do not.

I want a factory that can change direction and still keep the sequence, the material and the people who already know the work. That is what I mean by a responsive factory.