How to Improve Safety in Smart Manufacturing Operations

Smart manufacturing can make factories faster, more connected and more automated. But there is one area where “faster” should never mean “take shortcuts”: worker safety.

A modern production line may include robots, automated conveyors, vision systems, autonomous equipment, connected sensors, PLCs and remote monitoring. These technologies can reduce some traditional risks, but they can also introduce new ones.

For example, an operator may no longer manually lift a heavy component because a robot handles it. That improves ergonomics. But maintenance personnel may now need to work around equipment capable of moving automatically or receiving commands through a control system.

The goal should therefore be:

Use smart technology to strengthen existing safety principles—not replace them.

A useful approach is:

Key Safety Practices and Standards

1. Start With Physical Hazards

It is easy to become distracted by software, AI and dashboards.

The first question should still be:

What can physically hurt someone?

Look for hazards involving:

• rotating equipment;
• conveyors;
• pinch points;
• cutting tools;
• presses;
• robots;
• automated movement;
• electrical systems;
• pneumatic and hydraulic energy;
• heat;
• chemicals;
• stored energy.

OSHA notes that moving machine parts can cause serious injuries and that machinery capable of causing injury needs appropriate safeguarding.

Smart sensors can monitor equipment, but a dashboard cannot substitute for proper machine guarding.

2. Never Bypass Machine Guards for Productivity

Production pressure can create dangerous habits.

A guard is inconvenient.

An interlock causes the machine to stop.

A sensor makes setup slightly slower.

Someone decides to bypass it.

That may save seconds during a cycle while creating a much larger risk.

Guards and protective devices should be designed so operators can perform their normal work without unnecessary interference. OSHA specifically notes that safeguards should protect workers without creating new hazards or making normal work unnecessarily difficult.

If employees frequently want to bypass a safety device, investigate why.

The correct solution may be improving the machine design or workflow—not removing the protection.

3. Control Hazardous Energy During Maintenance

Smart machines may stop automatically, but stopped does not necessarily mean safe.

Equipment may still contain:

• electrical energy;
• compressed air;
• hydraulic pressure;
• gravity;
• stored mechanical energy;
• heat.

A machine could also restart unexpectedly.

In the United States, OSHA’s lockout/tagout requirements address hazardous energy during servicing and maintenance. They require appropriate energy-control procedures, training and periodic inspections.

A practical maintenance sequence is:

Connected software should never give maintenance workers a false sense that pressing “Stop” on an HMI is equivalent to energy isolation.

4. Make Robot Safety Part of the Complete Cell

Industrial robots deserve special attention because the robot itself is only part of the risk.

Consider an automated cell containing:

Robot + Gripper + Conveyor + Fixture + Workpiece

A safe robot does not automatically make the entire application safe.

Hazards can come from:

• robot movement;
• end effectors;
• sharp workpieces;
• welding operations;
• dropped components;
• surrounding machinery;
• unexpected movement during maintenance.

ISO 10218-1:2025 covers safety requirements for industrial robots themselves, while ISO 10218-2:2025 addresses industrial robot applications and cells, including integration, commissioning, operation and maintenance.

This distinction matters.

Evaluate the whole automated application, not just the robot specification.

5. Use Smart Sensors to Detect Unsafe Conditions

Connected technology can add another useful layer of safety monitoring.

Depending on the process, manufacturers may monitor:

• temperature;
• pressure;
• vibration;
• gas concentration;
• machine speed;
• door or guard status;
• equipment condition;
• environmental conditions.

For example, abnormal bearing temperature may indicate developing equipment trouble. Detecting the condition early can allow maintenance to investigate before a more serious failure occurs.

But sensor-based safety monitoring should be properly engineered and validated where it performs a safety function.

A general-purpose IIoT sensor or dashboard should not automatically be treated as a certified safety system.

6. Control Alarm Overload

More alarms do not automatically create a safer factory.

Imagine an operator receiving dozens of messages:

Motor temperature high

Network warning

Sensor disconnected

Pressure warning

Production target missed

Eventually, important alarms can disappear inside the noise.

Separate notifications into meaningful levels.

Every important alarm should answer:

What happened?

Who must respond?

What should they do?

When should it be escalated?

7. Include Cybersecurity in Safety Planning

In a traditional factory, safety and cybersecurity could sometimes be treated as separate topics.

Smart manufacturing makes that increasingly difficult.

PLCs, SCADA systems, robots, engineering workstations and other operational technologies may be connected through industrial networks.

NIST SP 800-82 Rev. 3 specifically recommends protecting OT while accounting for its unique performance, reliability and safety requirements.

Important practices include:

• OT asset inventory;
• network segmentation;
• controlled remote access;
• individual user accounts;
• backups;
• configuration management;
• monitoring;
• incident-response procedures.

A poorly controlled software or network change can affect a physical process. That is why cybersecurity should be part of operational risk management.

8. Train People for Normal and Abnormal Conditions

Operators should know more than which button starts the machine.

Training should cover:

• machine hazards;
• protective devices;
• emergency stops;
• alarm response;
• abnormal machine behavior;
• safe restart procedures;
• maintenance restrictions;
• reporting damaged safeguards.

OSHA emphasizes that even sophisticated safeguards are ineffective if workers do not understand their purpose and correct use.

Training becomes especially important after:

• installing new automation;
• modifying machine logic;
• introducing robots;
• changing guards;
• changing production processes.

People need to understand what changed—not just receive a new work instruction.

9. Use Data to Learn From Near Misses

One of the useful advantages of connected manufacturing is that operational events can often be reconstructed.

Suppose a conveyor unexpectedly stopped while an operator was nearby.

Instead of simply restarting production, engineers may review:

• PLC events;
• alarms;
• sensor states;
• timestamps;
• machine speeds;
• operator reports.

The goal should not be to find someone to blame.

The useful question is:

What combination of conditions allowed this situation to occur?

Near misses, recurring alarms and unsafe observations can provide valuable information before an actual injury happens.

10. Make Safety Part of Continuous Improvement

Safety should not be reviewed only after an incident.

ISO 45001 provides an occupational health and safety management framework built around hazard identification, risk assessment, worker participation, incident investigation and continual improvement.

A smart manufacturing safety cycle can follow the same principle:

Whenever a machine, robot, program or production process changes, ask:

Has the risk changed too?

That question should become part of normal engineering practice.

Common Smart Manufacturing Safety Mistakes

Avoid:

• assuming automation automatically makes work safe;
• bypassing guards or interlocks;
• treating an HMI stop as energy isolation;
• assessing a robot without assessing the complete cell;
• generating too many meaningless alarms;
• giving vendors uncontrolled remote access;
• making PLC or network changes without considering safety impact;
• ignoring operator feedback;
• failing to reassess risk after process changes.

One lesson deserves repeating:

A more automated factory still needs disciplined safety management.

In many cases, it needs even more.

Conclusion

Improving safety in smart manufacturing is not about adding another safety dashboard.

It requires combining traditional industrial safety with modern connected technology.

A practical approach is:

Robots can remove people from dangerous jobs.

Sensors can identify abnormal conditions.

Data can help engineers investigate incidents.

Automation can reduce repetitive manual work.

But none of these technologies eliminates the need for proper machine guarding, energy control, risk assessment and trained people.

The smartest factory is not simply the one with the most automation.

It is the one that uses automation and data while keeping safe operation at the center of every production decision.