Machine Design #08: Automating Screw Tightening – It's Not Just Mounting a Driver on a Robot

Screw tightening is a very familiar step in manufacturing.

From the outside it looks simple: place the screw in the hole, tighten it with a driver, and it is done. But in mass production, this step easily produces many faults:

A forgotten screw.
Undertightening.
Overtightening that strips the thread or cracks the plastic.
A crooked or tilted screw.
A screw with its head standing proud.
A screw that does not engage the thread, yet the machine reports it as tightened.
Every worker tightens in their own way, so quality is unstable.

So when we talk about automating screw tightening, the question should not only be:

"Which robot and which driver should I use to tighten the screws?"

The better question is:

"How do I get screws fed stably, the product positioned correctly, the tightening torque controlled, faults detected, and the tightening data traceable when needed?"

This article shares a practical way of thinking when you want to automate the screw-tightening step in manufacturing.

1. Before automating, understand where the current faults are

Many people who want to automate screw tightening think immediately of a robot.

But in reality, the robot is only one part of the system. If the product is placed off-position, the screw feed is unstable, the fixture is weak, or the screw type is not suited to automatic feeding, then no matter how accurate the robot is, it is hard to produce stable quality.

Before choosing a machine, check the following points:

What kind of fault is the current screw-tightening fault?
Is the fault due to the operator or due to the product design?
Is the screw easy to feed automatically?
Is the screw-hole position stable?
Does the product bend, warp, or deform when tightened?
Do you need torque management?
Do you need to store tightening data per product?
Does the product have many models?
What is the target cycle time?

If you buy a machine before clarifying these points, you can easily end up with a situation where:

the machine tightens, but production is unstable.

2. What benefits does automating screw tightening bring?

The biggest benefit of automating screw tightening is stable quality.

When done by hand, quality depends heavily on the worker's skill and condition. An experienced person can sense whether the screw has engaged the thread and whether the tightening force is abnormal. But when the task is repeated hundreds or thousands of times a day, mistakes are hard to avoid.

Automation reduces those deviations.

A good screw-tightening system can control:

The tightening torque.
The rotation speed.
The rotation angle.
The tightening time.
The downward travel of the driver head.
The OK/NG result.
The data for each screw position.

In addition, automation helps to:

Reduce dependence on worker skill.
Reduce missing-screw and loose-screw faults.
Increase production speed.
Reduce simple, repetitive actions.
Reduce worker fatigue.
Make it easy to build traceability for the product.

Especially for electrical products, mechanical equipment, safety-critical products, or products that need fault tracing, storing screw-tightening data is a very important point.

3. But full automation is not always the right answer

Automation is not always the best solution.

If the product still changes constantly, the quantity is small, or the screw positions change a lot from model to model, investing in a fully automatic screw-tightening machine may not be effective.

The reason is that a dedicated machine usually needs:

A dedicated fixture design.
Dedicated programming.
Feeder adjustment for the screw type.
Re-teaching the tightening positions when the model changes.
Rechecking the torque conditions.
Maintenance and fault handling when a screw jams.

If the volume is not yet large enough, or the product is not yet stable, it is better to start with semi-automation first.

This is a point many factories easily overlook. Robotizing is not automatically good. What matters is choosing the right level of automation.

4. The levels of screw-tightening automation

Screw-tightening automation can be divided into four main levels.

4.1. Semi-automation with an electric driver and a screw feeder

This is the easiest way to start.

The worker still holds the electric driver to work, but screws are fed automatically from a screw feeder. The worker no longer has to pick up each screw by hand.

This suits when:

The volume is not too large.
The product has many models.
You want to reduce the screw-picking action.
You want to reduce dropped-screw and missing-screw faults.
You do not yet want to invest in a fully automatic machine.
The step still needs a human for flexible handling.

If you also add a driver-suspension arm or a spring balancer, the worker does not have to hold the full weight of the driver. The driver is also easier to keep perpendicular, which helps reduce tilted-tightening faults.

In many cases, just doing this semi-automation level well improves the shop floor a great deal.

Do not view semi-automation as a "low-grade" option. For multi-model production, it is sometimes the most reasonable option.

4.2. A single-axis screw-tightening machine

A single-axis screw-tightening machine uses one driver unit to tighten each screw in order.

The driver unit can be mounted on:

An X-Y stage.
A Cartesian robot.
A SCARA robot.
A dedicated mechanism.
A servo slide table.

This option suits when:

One product has a few screw positions.
The screw positions are on the same plane.
The product is relatively stable.
You need automation but still want to keep flexibility.
The cycle time is not too tight.

The advantage is that the program is easier to change than a multi-axis machine. When the model changes, you can change the fixture and adjust the tightening coordinates.

The disadvantage is that, because it tightens one screw at a time, the cycle time is longer than tightening many screws at once.

4.3. A multi-axis screw-tightening machine

A multi-axis screw-tightening machine uses several drivers to tighten many screws at once.

For example, if a product has 4 screws at 4 corners, the machine can use 4 drivers to tighten them simultaneously. With a good design, the cycle time will be very short.

This option suits when:

The volume is large.
The product is stable for the long term.
The screw positions change little.
Cycle time is an important requirement.
Continuous production is needed.

The advantage is high speed.

The disadvantage is low flexibility. If the product changes the hole positions, the size, or the number of screws, the machine may need quite a lot of mechanical rework.

Simply put:

A stable product with a large volume → a multi-axis machine is advantageous. Many models that change frequently → weigh it carefully.

4.4. A screw-tightening robot

A screw-tightening robot is the option of mounting the screw-tightening driver on a robot.

You can use:

A Cartesian robot.
A SCARA robot.
A 6-axis robot.
A collaborative robot.

A SCARA suits screw tightening on a plane at high speed, often used in electronics assembly or small mechanical units.

A 6-axis robot suits when you need to tighten at various angles, or when the product has a complex shape.

A collaborative robot suits when you want to place it near people, save space, or combine a human worker and a robot in the same cell.

However, do not simply assume a collaborative robot is always safe and needs no risk assessment. When the robot head carries a driver, bit, Z-axis, and screw-feed mechanism, you still need to consider speed, force, pinch points, and the operator's actions.

A robot is more flexible than a dedicated machine, but a screw-tightening robot system still depends heavily on:

The fixture.
The feeder.
The driver.
The inspection sensors.
The model-management program.
The safety design.

The robot cannot solve everything if the underlying mechanics are not stable.

5. How does a screw-tightening machine differ from a screw-tightening robot?

These two concepts are easy to confuse.

A screw-tightening machine is usually a device dedicated to the screw-tightening step. It can include the driver, feeder, Z-axis, fixture, and a motion mechanism.

A screw-tightening robot is a system that uses a robot as the motion platform, then mounts the driver and related equipment to perform the tightening.

Criterion	Dedicated screw-tightening machine	Screw-tightening robot
Goal	Optimized for one product/step	Flexible for many positions, many models
Speed	Can be very fast	Depends on the robot and the path
Flexibility	Low to medium	Higher
Best fit	Large volume, few model changes	Many models, many positions
Cost of change	Can be high if the product changes	Easier if well designed
Complexity	Much dedicated mechanics	More complex programming and safety

No option is always better. Whether the choice is right or wrong depends on the product and the production conditions.

6. What does an automatic screw-tightening system consist of?

An automatic screw-tightening system usually has four main parts:

The tightening driver.
The Z-axis.
The screw feeder.
The controller.

In addition, you need a fixture, sensors, safety mechanisms, and a robot or motion axis if you want full automation.

6.1. The tightening driver

The driver is the part that directly rotates and tightens the screw.

For a simple step, an ordinary electric driver can be used. But if quality management is needed, choose a driver that can control torque, speed, and angle, and can output data.

Parameters to watch:

The torque range.
The torque accuracy.
The rotation speed.
Rotation-angle control.
Fault-detection ability.
Communication with a PLC or PC.
The ability to store tightening data.

Do not choose a driver by maximum torque alone. Choose by the actual working torque range and the stability at that range.

6.2. The Z-axis

The Z-axis brings the driver down to contact the screw and the product.

You can use:

A pneumatic cylinder.
A servo axis.
A spring mechanism to assist the pressing force.
A dedicated Z module for screw tightening.

A pneumatic cylinder is simple and cheap, but it is hard to control the descent speed and pressing force accurately.

A servo Z allows better control of:

The descent speed.
The stop position.
The pressing force.
The motion profile.
The travel after the screw starts to engage the thread.

In automatic screw tightening, the downward pressing force is very important. Too little force and the bit slips easily. Too much force and it easily deforms the product, damages the thread, or makes the screw engage wrongly.

6.3. The screw feeder

The screw feeder is a part that is easily underestimated, yet it largely determines the stability of the system.

The feeder's job is to bring screws from a loose state to one at a time, in the right orientation, at the right moment.

Some common types:

Bowl feeder.
Rail-type feeder.
Blow feeding through a tube.
Feeding to a pick-up position.
Feeding directly to the driver head.

When choosing a feeder, test it with the actual screw type. Do not rely on the catalog alone.

Factors affecting feedability:

The screw diameter.
The screw length.
The head shape.
The length-to-diameter ratio.
The material and surface treatment.
Whether the screw is oily.
Whether the screws easily overlap each other.
Whether the screws easily jam in the rail or tube.

Many unstable systems are unstable not because the robot is poor, but because the screw feed is unstable.

6.4. The controller

The controller is the part that controls and monitors the system.

A good controller not only commands the tightening but must also detect faults.

For example:

No screw.
A wrongly fed screw.
Torque NG.
Angle NG.
An abnormal tightening time.
A stripped thread.
A proud screw.
The driver not returned to home.
A screw jammed in the feeder.
The product not placed in the correct position.

If the machine only tightens but does not detect faults, then automation can make faults happen in large numbers, faster.

This is a very important point.

7. The fixture matters no less than the robot

In screw-tightening automation, the fixture is the foundation.

If the product sits in the fixture unstably, the screw-hole position shifts, the product bends, or it gets lifted when the driver comes down, then the tightening conditions will not be stable.

A good fixture must ensure:

Firm positioning of the product.
A clear reference surface.
No deformation of the product.
Enough clearance for the driver head.
A mechanism to prevent placing it the wrong way round.
A sensor confirming the product is placed correctly.
Easy loading and unloading of the product.
Easy cleaning and maintenance.

A common mistake is trying to use a camera to compensate for a poor fixture.

A camera can compensate for small deviations, but you should not use a camera to fix an unstable mechanical design.

A better principle is:

Make the fixture stable first. Use the camera only to compensate for the remaining deviation.

8. What should you inspect after tightening?

Finishing the tightening does not mean it passed.

Depending on the quality requirement, you may need to inspect:

Whether the number of screws is complete.
Whether a screw is standing proud.
Whether the torque is within the limits.
Whether the angle is abnormal.
Whether the screw depth is correct.
Whether there is a sign of a stripped thread.
Whether a position was tightened by mistake.
Whether it is the correct screw type.

Inspection options can include:

Torque/angle data from the driver.
A sensor checking the screw-head height.
A camera checking screw presence/absence.
A laser displacement sensor.
A mechanical inspection mechanism.
Linking the data to the product's serial code.

For important products, store the data per product. When a fault appears in the field, this data helps trace back the production process.

9. Automating bolt and nut tightening

Not only small screws — bolts and nuts can also be automated.

For applications that need large torque and high reliability, an electric nutrunner is usually used. A nutrunner can control torque and angle and store data better than simple tools.

When automating bolts/nuts, also watch for:

The reaction force during tightening is larger.
The fixture must be rigid enough.
The tool can be heavier.
Safety is more important.
Feeding bolts/nuts needs its own orientation mechanism.
A tool changer may be needed if there are many types of bolt/nut.

Bolts and nuts are usually easier to grip than small screws because of their larger size, but the requirements for tightening force, rigidity, and safety are higher.

10. Which option should you choose?

You can choose along the following lines.

Production condition	Option to consider
Low volume, many models	Semi-automation: electric driver, feeder, suspension arm
Medium volume, stable screw positions	Single-axis machine, Cartesian robot, SCARA
Large volume, stable product	Multi-axis machine, dedicated machine
Many models, need flexibility	Screw-tightening robot
Traceability needed	Driver/nutrunner with data output
Screws hard to feed automatically	Test the feeder before designing the machine
Product deforms easily	Prioritize the fixture and pressing-force control

The key point is not to choose equipment by feel.

Start from:

The product.
The screw type.
The number of tightening positions.
The cycle time.
The quality requirement.
The degree of model change.
The installation space.
The budget.
The maintainability on-site.

11. On the investment cost

The cost of screw-tightening automation varies enormously.

A semi-automation option may consist only of:

An electric driver.
A screw feeder.
A suspension arm.
A simple fixture.

But a fully automatic system can consist of:

A robot.
A servo driver.
A feeder.
A Z-axis.
A fixture.
Sensors.
A camera.
A PLC.
Safety equipment.
A data-collection PC.
Traceability software.

When calculating the cost, do not look only at the price of the main equipment.

Also account for:

The mechanical design.
Machining the fixture.
The electrical design.
PLC/robot programming.
Adjusting the tightening conditions.
A trial run with the real product.
Instruction documentation.
Operator training.
Spare parts.
The time for improvement after entering production.

A cheap system that runs unstably will cost far more than a system designed correctly from the start.

12. Some makers you may encounter

In the field of automatic screw tightening, there are many makers depending on the purpose.

Some names commonly seen in Japan:

Nitto Seiko: screws, screw-tightening machines, automatic assembly systems.
ESTIC: electric nutrunners, torque/angle control.
HIOS: electric drivers, small-screw applications.
Janome: bench robots, automatic screw-tightening applications.
Daihen: industrial robots, robot systems with tool changers.
Uryu, Yokota: industrial tightening tools.

However, when choosing a maker, do not only ask "how much does the machine cost?"

Prepare the following information in advance:

The product drawing.
The screw positions.
The screw type.
The required torque.
The production quantity per day.
The target cycle time.
The number of models.
The OK/NG requirement.
The data-storage requirement.
The installation space.
How the product is loaded and unloaded.
The safety requirement.

The clearer the information, the more realistic the proposal you will receive.

13. A checklist before deploying screw-tightening automation

Before ordering a machine or requesting a quote, check the following points:

What is the goal of automation?
Where are the current faults happening?
Is the screw type suitable for automatic feeding?
Is the screw-hole position stable?
Does the product deform when tightened?
Is the fixture stable enough?
Do you need to check for proud screws?
Do you need torque/angle management?
Do you need to store data by product serial?
Does the product change model frequently?
What is the target cycle time?
Is there enough installation space?
Which part will the worker handle?
When the machine gives an NG, how is the product handled?
When the feeder jams a screw, who handles it?
Are replacement driver bits easy to buy?
Do you need a margin for future design changes?

If you cannot yet answer these questions, do not rush to finalize the equipment option.

Conclusion

Automating screw tightening is not just mounting a driver on a robot.

A good system must handle the whole process:

stable screw feeding → correct product positioning → bringing the driver to the right position → controlling the tightening force → detecting faults → storing data → ensuring safety and easy operation.

If the product still changes a lot, start from semi-automation: an electric driver, a screw feeder, a suspension arm, and a good fixture.

If the product is stable, the volume is large, and there are clear quality requirements, only then should you consider an automatic screw-tightening machine or a screw-tightening robot.

Automation should not start from the equipment. It should start from the shop floor: what fault the product currently has, which action wastes time, what data needs to be managed, and whether that step is truly suitable for automation.

That is the way to deploy sustainable screw-tightening automation in manufacturing.

A note from MINATA

MINATA shares these notes as reference material from a practical machine-design point of view. If you would like to discuss automating a tightening step, confirm drawings, compare options, or interpret Japanese–Vietnamese engineering terms during design, feel free to get in touch and we can review it together.

View all MINATA technical articles