Why Do Humans Become Busier in Increasingly Automated Factories?

Subtitle
On the invisible labor that increases behind the scenes of labor-saving

Introduction

In recent years, many companies have been promoting DX.

DX is not simply about digitizing paper.
It is about reconstructing tasks that people used to handle based on experience and intuition using data and systems, thereby changing the entire flow of operations.

DX in factories is the same.

Displaying work instructions on a terminal.
Reading part numbers with QR codes.
Recording the usage history of parts.
Sharing information on out-of-stock items.
Transporting with AGVs.
Preventing picking errors with error-proofing lamps.

All of these are introduced to reduce the burden on humans, stabilize quality, and keep production from stopping.

In the background, there is also a labor shortage.

There are not enough people.

Skilled workers are decreasing.

It takes time to teach everything to new employees.

The workplace has become too complex to rely solely on human attention.

That is why companies try to support the workplace through DX and automation.

That in itself is not wrong.

Rather, it is an unavoidable trend for factories from now on.

However, the problem is that promoting DX and automation does not necessarily mean that humans will have an easier time.

If DX is about reconstructing the flow of information, then automation is about reconstructing the flow of goods and work.

In factories, these two are progressing in an overlapping manner.

Terminals move information, AGVs move goods, and error-proofing lamps guide human judgment.

In other words, modern factories are attempting to reorganize the flow of people, goods, and information through systems.

Automation is progressing in factories.

AGVs run.

Error-proofing lamps light up.

When a QR code is scanned, the next task begins.

Out-of-stock information is displayed on the terminal.

Critical safety parts are managed for traceability on a separate terminal.

At first glance, it looks like a rational scene.

People don't have to carry things.

People don't have to search.

People don't have to memorize.

People don't have to make judgments.

Thinking this way, automation seems to be in the right direction.

However, when you stand on the shop floor, you see a slightly different view.

AGVs are running.

But humans perform the battery replacements.

Error-proofing lamps indicate parts.

But those lamps break frequently.

Scanning a QR code is the signal to start work.

But if the barcode reader or the product label is dirty, the reading becomes unstable, and it is possible that the wrong part number is displayed.

There is a terminal to convey out-of-stock information.

There is a terminal to record critical safety parts.

There is a terminal to display part numbers.

There are error-proofing lamps.

There are scanners.

The functions are increasing.

However, they are not unified.

As a result, instead of having free time, humans actually become busier.

Why?

It is because automation did not make human work disappear, but rather moved it to places where it is harder to see.

Automation does not eliminate work; it changes the location of work.

In old factories, humans carried, humans searched, humans memorized, and humans made mistakes.

So companies thought.

Mistakes happen because we rely on human judgment.

Then, let's have machines give instructions.

The burden is heavy because humans carry things.

Then, let's have AGVs carry them.

Omissions occur because humans record them.

Then, let's have terminals record them.

This way of thinking itself is not wrong.

Rather, it is a necessary idea for stabilizing factory quality.

However, there is a problem with automation that is easily overlooked.

Introducing machines does not mean human work disappears.

In many cases, human work shifts from operation to recovery.

From transport to monitoring.

From memory to verification.

From judgment to cross-checking.

From simple tasks to exception handling.

In other words, humans become interpreters of the system rather than operators.

Herein lies the reason why humans become busier in automated factories.

AGVs run automatically, but they are not self-contained.

When AGVs are running in a factory, it looks like automation is progressing.

People don't have to push carts.

They run on set routes.

Transport can be standardized.

The burden of heavy objects is also reduced.

Indeed, that is a major improvement.

But what if those AGVs stop frequently?

Humans perform battery replacements.

Humans watch for sensor errors.

Humans respond when traffic jams occur.

Humans recover stopped AGVs.

At this time, the AGV is not completely eliminating human work.

It is changing human work from transport to caretaking.

Of course, this is not to say that AGVs are bad.

The problem is whether the work reduced by introducing AGVs and the work increased to maintain the AGVs are being properly compared.

Transport time may have decreased.

But has recovery time increased?

The movement of operators may have decreased.

But have calls for maintenance staff increased?

The burden of humans carrying things may have decreased.

But has the burden of humans having to accommodate the needs of machines increased?

If you don't look at this, the evaluation of automation will be one-sided.

The more error-proofing there is, the more new errors are born.

The same goes for error-proofing lamps.

To prevent picking the wrong parts, a lamp lights up.

The operator takes the part from the lit location.

After taking it, they press a button.

The system confirms that the work has progressed.

As a mechanism, it is very easy to understand.

But if those terminals cost over 10,000 yen each, and there are hundreds of them, and they break frequently, the story changes.

Error-proofing exists to prevent mistakes.

But if the error-proofing itself breaks, it becomes a source of new anxiety.

Does this lamp really light up?
Is this button responding?
Is this display correct?
Is it a communication error?
Is it a terminal failure?
Is it a connector contact failure?
Is it an address issue?

Operators and maintenance staff start to doubt the mechanism itself, not the parts.

For example,

Suppose that because three error-proofing units are acting up today, you have to check the work instructions.

This is a quite ironic state.

The machine introduced to prevent human error is demanding new judgments from humans.

For the sake of error-proofing, even more error-proofing is needed.

By this point, the shop floor doesn't know if it is working or maintaining the system.

When terminals increase, correctness becomes fragmented.

Another big issue is that functions are not unified.

There is a terminal for scanning QR codes.

There is a terminal for conveying out-of-stock information.

There is a traceability terminal for critical safety parts.

There are error-proofing lamps.

There is a screen displaying part numbers.

Each purpose is correct.

QR codes are necessary to start work.

Out-of-stock information is necessary to keep the shop floor from stopping.

Traceability for critical parts related to running, turning, and stopping is necessary for quality assurance.

Error-proofing lamps are necessary to prevent picking errors.

Each one is correct.

However, when each runs on a separate terminal, it is only integrated in the operator's head.

This is the problem.

The operator must think as follows.

Does the QR part number match the displayed part number?
Is the out-of-stock information related to this task?
Is the recording of critical parts complete?
Does the lamp instruction match the current part number?
When information between terminals conflicts, which one should I believe?

This is no longer simple work.

It is information cross-checking work.

In other words, even though the factory as a whole is digitizing, only the final integration process is being performed inside the human brain.

Herein lies the contradiction of the modern factory.

Just because the barcode reader is dirty, the world goes out of sync.

QR codes and barcodes are very convenient.

You know the part number just by reading it.
Work progresses.
Records remain.
Human input errors also decrease.

But what happens if the scanner or the product label is dirty?

It cannot be read correctly.
Reading becomes unstable.
The display is delayed.
In some cases, the wrong part number appears.
Or, it looks like the previous information remains.

What is important here is not the fact that the machine makes a mistake.

It is whether the design allows humans to notice that mistake when the machine makes one.

If it couldn't read it, it should clearly state that it cannot read it.

If it doesn't match, it should make it impossible to start work.

If it is uncertain, it should indicate that it is uncertain.

But on the shop floor, uncertain things are often displayed with a face that looks certain.

This is dangerous.

The machine introduced to reduce work relying on human intuition now demands that humans use intuition to doubt the machine's display.

At this moment, automation is not making humans' lives easier.

It is demanding that humans doubt in a more sophisticated way.

Management of critical safety parts is necessary, but the burden on the shop floor is also a reality.

Parts related to running, turning, and stopping are managed particularly strictly.

This is natural.

Whether it is a car or a machine, parts related to movement or stopping are directly linked to safety.

That is why traceability is needed.
Who, when, which part, and in which product was it used?
It must be traceable when a defect occurs.

This mechanism cannot be ignored from the perspective of quality assurance.

However, from the perspective of the shop floor, this is also additional work.

Perform normal work.
Read the QR code.
Look at out-of-stock information.
Look at the error-proofing lamps.
If it is a critical part, record it on a separate terminal.
Check if there are any recording omissions.

Even if it is work necessary for safety, it remains a burden on the shop floor.

Therefore, companies must look not only at the necessity but also at the form of the burden.

Just saying 'do it because it's important' is not enough.

Precisely because it is important, it must be designed to be hard to make mistakes, hard to get confused, and processable in a single flow.

The real problem is not that there are more machines

I must not be misunderstood here; this is not about machines or terminals being bad.

AGVs are necessary.
Poka-yoke lamps are necessary.
QR codes are necessary.
Out-of-stock information is necessary.
Traceability of critical parts is necessary.

The problem is that they each hold their own version of 'correctness.'

Correctness of part numbers
Correctness of stock status
Correctness of work start
Correctness of part instructions
Correctness of record completion

When these exist in separate places, workers on the floor must reconcile multiple versions of correctness.

What is truly needed in a factory is not multiple versions of correctness.

It is a single flow that workers can follow without hesitation.

What should I do now?
Is this the right part?
Are there any items out of stock?
Is a record required?
Is the record complete?
Can I proceed to the next step?

That is all the worker wants to know.

However, when functions are fragmented, the worker becomes a manager of information.

This is where humans become busy.

It is not that their hands are busy.
It is that their minds are busy.

The invisible labor created by automation

As automation progresses, visible work decreases.

Fewer people are seen pushing carts.
Fewer people are seen searching for parts.
Fewer people are seen recording on paper.

Instead, invisible work increases.

Looking at terminals
Reconciling information
Suspecting anomalies
Wiping down readers
Checking for lamp malfunctions
Replacing AGV batteries
Restoring stopped equipment
Judging the meaning of displays
Bridging discrepancies between multiple systems

These are not tasks that directly produce products.

But if these are not done, products cannot be made.

In other words, in modern factories, work to maintain the state where products can be made is increasing more than the work to make the products themselves.

This is hard to see.

Documents for equipment introduction list reductions in transport time and work errors.

However, recovery time, confirmation fatigue, the load of reconciling between terminals, the frequency of replacing maintenance parts, and confusion on the floor are hard to see.

That is why they are difficult to improve.

Invisible labor is easily treated as if it does not exist.

But the busyness on the floor is exactly where it lies.

Automation as a corporate philosophy

What a company needs is not whether it has automated.

It is whether human hesitation has decreased.

It is not whether it has introduced AGVs.
It is whether the load on the floor, including taking care of the AGVs, has decreased.

It is not whether it has installed Poka-yoke lamps.
It is whether the Poka-yoke itself has become a new cause of anxiety.

It is not whether it uses QR codes.
It is whether uncertainty is correctly communicated when reading is uncertain.

It is not whether there are traceability terminals.
It is whether recording is naturally integrated into the work flow.

It is not whether there are terminals.
It is whether the floor can use them without hesitation.

Here, corporate philosophy is questioned.

Increasing machines is relatively easy.

Functions can be added.
Terminals can be placed.
Rules can be increased.
Check items can be increased.

However, if they do not reduce human hesitation, the floor will not become easier.

Rather, the work to maintain correctness increases.

What a company should really think about is not the number of equipment.

It is where humans are hesitating.

Humans are not the final integration device

Unintegrated systems eventually push everything onto humans.

The convenience of AGVs
The convenience of Poka-yoke lamps
The convenience of QR codes
The convenience of out-of-stock terminals
The convenience of traceability terminals
The convenience of production planning
The convenience of quality assurance

Each convenience gathers on the people on the floor.

And humans interpret, reconcile, judge, and continue working on the spot.

Humans are flexible.

That is precisely why companies rely on humans.

Even if systems are somewhat fragmented, humans manage.
Even if displays are hard to understand, humans memorize them.
Even if there are multiple terminals, humans memorize the order.
Even if anomalies occur, humans judge based on experience.

However, that is not design in the true sense.

It is just making humans bear the incompleteness of the system.

Humans are not the final integration device.

Humans are not there to absorb contradictions between machines.

Originally, humans should be enabled to concentrate on higher-value judgments.

What is true automation?

True automation is not about removing humans from the floor.

It is about creating a structure where humans do not have to hesitate.

A factory with many machines is not necessarily a good factory.
A factory with many terminals is not necessarily an advanced factory.
A factory where AGVs run is not necessarily an automated factory.

A truly automated factory is one where the human judgment load is reduced.

It is clear which information to look at.
It is clear which terminal to trust.
It is clear what to do in case of an anomaly.
Recording is naturally integrated into the work flow.
Uncertain information is displayed as uncertain.
Even the maintenance load of machines is designed in advance.

Including all this is automation.

Introducing equipment is just the entrance to automation.

True automation is designing how well humans can respond without hesitation when that equipment breaks, gets dirty, stops, or when information conflicts.

Conclusion

The reason humans become busy in increasingly automated factories is simple.

It is not that human work has disappeared.

It is just that human work has become harder to see.

The work of transporting has decreased.
However, the work of restoring stopped machines has increased.

The work of searching has decreased.
However, the work of doubting the correctness of displays has increased.

The work of memorizing has decreased.
However, the work of reconciling information from multiple terminals has increased.

The work of recording has decreased.
However, the work of confirming whether recording is truly complete has increased.

Automation is not something that makes humans comfortable if left alone.

If the design is wrong, automation drives humans into the gaps of the system.

And humans become busy in those gaps.

That is why I think companies should ask these questions:

Have machines increased?

No.

Has human hesitation decreased?

Have terminals increased?

No.

Has the correctness of the floor been unified?

Has it been automated?

No.

Are humans not being chased by taking care of machines?

The future of a factory is not determined by the number of machines.

It is determined by how far human burden can be treated as something visible.

Making invisible labor visible.

That is true improvement, and I believe it is the philosophy a company should have.

Introduction to related articles 1

In this article, I considered why invisible human labor increases as factory automation progresses.

But the same structure is happening in stores.

Self-checkout, tablet ordering, serving robots, EC, smartphone payments.

When it looks like humans are disappearing from stores, are they really disappearing?

Or is human labor, judgment, and customer service just being reallocated to customers, machines, and data?

A related article that considered that problem is

What remains in a store where humans are disappearing?

If the factory article looked at automation from the production floor, this article looks at simplification from the consumption floor.

The question is the same for both.

Have humans disappeared?
Or have they been moved to a place that is hard to see?

Introduction to related articles 2

In this article, I considered how human work moves to hard-to-see places in increasingly automated factories.

AGVs run.

Poka-yoke lamps light up.

Part numbers are read with QR codes.

Out-of-stock information is displayed on terminals.

Critical safety parts are managed for traceability.

Each one is correct.

However, when they are left on the floor while fragmented, only the final reconciliation and judgment are pushed onto humans.

At that time, it is not just machine performance that is questioned.

Who is watching it?

Who is restoring it?

Who is absorbing the discrepancies in information?

Who is bearing the hesitation of the floor?

What is needed here is the manager's perspective.

Related article

A manager's job is to take responsibility.

In this related article, I think about the difference between a player and a manager not from work ability, but from the structure of responsibility.

Players perform the work in front of them sincerely.

Managers create a structure where that sincerity does not die in vain.

And when things don't work out no matter how hard they try, they don't run away at the end.

It is the same on an automated floor.

Just because you introduced machines doesn't mean it's over.

Just because you increased terminals doesn't mean it's an improvement.

Just because AGVs are running doesn't mean you've saved labor.

What you should really look at is not what the machines are doing.

It is who is bearing what the machines did not do.

Can you treat the invisible labor on the floor as something visible?

Are humans not being made into the final integration device?

Is the player's hard work not being consumed to fill the incompleteness of the system?

Seeing that is also a manager's responsibility.

If you are thinking about factory automation, talking about equipment is not enough.

How much do humans hesitate, confirm, restore, and judge in the gaps of that equipment?

If you don't look that far, it won't be a true improvement.

In that sense, this article and the manager theory deal with the same question.

Is the structure that is supposed to make humans comfortable not cornering them?

Is the responsibility of the floor not being pushed into invisible places?

The article that considered that question from the aspect of management responsibility is

A manager's job is to take responsibility.

is the article.