A safety light curtain looks simple: two aluminum bars facing each other across a machine opening.

That appearance is deceptive.

Behind those bars sits an active opto-electronic protective device, redundant safety outputs, diagnostic logic, a safety relay or safety PLC, downstream contactors or drives, and a calculated separation distance intended to buy enough time for hazardous motion to stop before a person reaches it.

Miss one link and the system can fail.

I have reviewed enough OSHA enforcement records to reach a blunt conclusion: most light-curtain failures are not mysterious sensor failures. They are integration failures, bypass decisions, bad safety-distance calculations, uncovered access routes, careless restart logic, or managers accepting a nuisance-trip “fix” that quietly removes the protection.

The technology works. The installation decides whether it protects anyone.

What a Safety Light Curtain Actually Does

A safety light curtain is a non-contact machine-guarding device that creates a detection field between an infrared emitter and a receiver. When a hand, arm, or body interrupts that field, the curtain changes its safety outputs and tells the machine control system to prevent or stop hazardous movement.

OSHA classifies light curtains as presence-sensing devices and explains that they are designed to stop a machine stroke when the sensing field is interrupted. The agency also warns that multiple technical conditions must be met before they can serve as point-of-operation safeguards.

A beam disappears.

The receiver detects the missing optical signal, drives its redundant Output Signal Switching Devices—normally called OSSD channels—to the OFF state, and forces the safety relay or safety PLC to remove permission for hazardous motion before the intruding person can reach the danger point.

Simple enough, right?

Not really. Detection is only the opening move. The machine still has to react, disconnect or control hazardous energy, stop within the validated time, remain stopped while the field is blocked, and avoid an unexpected restart after the person leaves.

That entire chain is the safety function.

For buyers comparing actual configurations, the site’s range of safety light curtains for machine guarding includes compact, general-use, heavy-machine, multi-sided, dual-output, ultra-thin, waterproof, and Type 4 options. Those categories address different installation problems; they should not be treated as interchangeable versions of the same sensor.

The Detection-to-Stop Sequence, Step by Step

1. The emitter creates the protective field

The emitter contains a vertical array of infrared transmitting elements. Rather than producing a solid wall, it sends multiple synchronized optical beams toward the receiver.

The beam arrangement defines several practical characteristics:

• Detection capability or resolution
• Protective height
• Operating range
• Alignment tolerance
• Resistance to optical interference
• Whether the field is suited to finger, hand, arm, or whole-body detection

A narrow gap between beams can detect smaller objects. A wider beam pitch may be acceptable for perimeter or body access, but it should not be casually substituted for point-of-operation hand protection.

OSHA is unusually direct here: perimeter light curtains with wider channel spacing are designed for perimeter or area safeguarding and cannot automatically be used as point-of-operation protection for fingers and hands.

2. The receiver continuously checks the beams

The receiver expects a specific optical pattern from the emitter. Under normal conditions, the required beams arrive correctly and the safety outputs remain in their permissive state.

When an opaque object interrupts enough of the field, the receiver identifies the intrusion. A safety-rated device also performs internal diagnostics intended to detect faults rather than simply reporting “beam clear” or “beam blocked.”

This is where buyers often confuse a safety light curtain with an automation light grid.

A standard optical array may count cartons, measure product height, detect an edge, or confirm that a component is present. That does not make it suitable for personnel protection. The distinction is examined in detail in safety light curtains versus non-safety light curtains.

My rule is harsh but useful: a sensor that can notice a box is not automatically a device I would trust with a hand.

3. The OSSD channels switch off

Most modern safety light curtains use two monitored semiconductor safety outputs, commonly labeled OSSD1 and OSSD2.

During normal operation, both channels indicate that the protective field is clear. When the field is interrupted or a relevant fault is detected, both channels transition to the safe state.

Why two channels?

Because a personnel-protection system should not depend on a single output transistor, a single wire, or a single control path behaving perfectly forever. The downstream controller checks the channels for expected behavior, discrepancies, shorts, cross faults, or timing problems according to the system architecture.

The curtain is not usually switching the machine motor directly. It is sending safety information to the next part of the circuit.

4. The safety controller evaluates the signal

The OSSD channels feed a safety relay, programmable safety controller, or safety PLC. That logic device evaluates the curtain state together with other inputs such as:

• Emergency-stop circuits
• Guard-door interlocks
• Two-hand controls
• Safety scanners
• Reset devices
• Contactor feedback
• Drive status
• Mode-selection signals

ISO 13849-1:2023 provides the current methodology for designing and integrating safety-related parts of control systems, including the hardware and software that perform safety functions. Importantly, it does not assign a required Performance Level to every machine; that requirement must come from the machine’s risk assessment and applicable machine-specific standards.

This distinction matters. A Type 4 curtain does not magically turn an ordinary control panel into a PL e safety system.

The sensor is one subsystem. The achieved performance depends on the complete architecture.

5. Final switching devices remove hazardous motion

After the safety controller receives the stop demand, it commands the machine’s final control elements.

Depending on the machine, that may involve:

• De-energizing redundant contactors
• Applying a clutch-brake stopping function
• Closing a monitored hydraulic valve
• Venting pneumatic pressure through a safety valve
• Activating Safe Torque Off on a variable-frequency drive
• Initiating a controlled stop before torque is removed

This is the point where brochure language meets mechanical reality.

A fast light curtain cannot rescue a slow machine. If a press, saw, robot axis, winding machine, or automated cutter needs 600 milliseconds to stop, the mounting distance must account for that total stopping behavior. Replacing a 20-millisecond sensor with a 10-millisecond sensor helps only at the edge; it does not erase the machine’s inertia.

6. Reset and restart logic keep the machine from moving unexpectedly

When the protective field becomes clear again, the machine should not restart merely because the person stepped back.

For many applications, a deliberate manual reset is required. The reset control should be positioned so the operator can verify that no one remains inside the safeguarded space, while being unable to actuate the reset from within that space.

External Device Monitoring, or EDM, can be used to confirm that downstream contactors or switching devices actually changed state. A welded contactor should not be allowed to masquerade as a successful stop.

This is one of the least glamorous parts of light-curtain engineering.

It is also one of the most important.

The Safety Chain Is Only as Strong as Its Slowest Element

Safety-chain element	What it does	Typical engineering question	Common failure
Emitter and receiver	Create and monitor the infrared field	Can the selected resolution detect the relevant body part?	Wrong beam spacing or poor alignment
OSSD outputs	Send redundant safety-state signals	Are both channels independently monitored?	Outputs wired into a standard PLC input
Safety relay or safety PLC	Evaluates the safety function	Does the architecture meet the required PL or SIL?	Single-channel logic or incorrect programming
Final control elements	Stop or inhibit hazardous motion	Can the contactors, valves, brake, or drive achieve the stop reliably?	Welded contactor, leaking valve, weak brake
Reset and EDM	Prevent unexpected restart and monitor switching	Can the operator inspect the protected space before reset?	Automatic restart or hidden reset button
Mechanical machine system	Physically decelerates and stops	What is the measured worst-case stopping time?	Using a catalog estimate instead of a stop-time test
Physical and supplemental guards	Block access outside the optical field	Are side, rear, over, under, and pass-through routes covered?	An employee walks around or remains behind the curtain

The hard truth is that the curtain itself may perform exactly as designed while the overall machine remains unsafe.

OSHA’s general machine-guarding rule, 29 CFR 1910.212, requires guarding against point-of-operation hazards, ingoing nip points, rotating parts, flying chips, and sparks. It recognizes electronic safety devices as one possible method, but it does not say one optical sensor solves every hazard.

A light curtain cannot contain a broken blade.

It cannot stop a metal fragment.

It cannot prevent someone from reaching around an uncovered side opening.

And it cannot compensate for a machine that cannot stop during the hazardous portion of its cycle.

Safety Distance: Where Most Installations Become Engineering Work

The curtain must be far enough from the hazard that the machine reaches a safe state before the person reaches the danger point.

The basic engineering relationship is often expressed conceptually as:

Separation distance = approach speed × total response time + additional penetration allowance

The total response time is not just the light curtain’s response time. It can include:

• Sensor response time
• Safety-controller response time
• Network or interface delay
• Output-device response time
• Valve, contactor, drive, clutch, or brake response
• Mechanical stopping time
• Brake-monitor allowance
• Additional delays introduced by filtering or logic

The current international positioning standard is ISO 13855:2024, published in November 2024. It replaced ISO 13855:2010 and covers the positioning and dimensioning of safeguards, including electro-sensitive protective equipment such as active opto-electronic protective devices.

That date matters. Specifications still referring only to ISO 13855:2010 are pointing to a withdrawn edition.

In the United States, OSHA’s mechanical-power-press guidance also shows why total response matters. Its safety-distance formula accounts for press stopping time, control response, presence-sensing-device response, brake-monitor allowance, and penetration depth. OSHA uses a hand-speed constant of 63 inches per second in that specific mechanical-power-press context and requires actual stop-time measurement rather than guesswork.

Do not copy that value blindly into every machine application.

Different machines, jurisdictions, standards, approach directions, detection capabilities, and installation geometries can require different calculations. The formula must match the applicable standard and machine type.

A practical example

Suppose a system has the following measured or documented times:

• Light curtain response: 15 ms
• Safety-controller response: 10 ms
• Output and drive response: 25 ms
• Mechanical stopping time: 320 ms

The preliminary total is 370 ms, before adding any required tolerances, brake-wear allowance, penetration factor, or application-specific margin.

At an assumed approach speed of 1,600 mm/s, 370 ms alone represents 592 mm of travel.

That is why mounting a curtain 200 mm from a hazard because “the sensor reacts in 15 milliseconds” is not engineering. It is arithmetic malpractice.

Resolution, Type, and Protective Height Are Different Decisions

People routinely mix these terms together.

They should not.

Detection resolution

Resolution describes the smallest object the protective field can reliably detect under defined conditions. It influences whether the application is intended to detect fingers, hands, arms, or bodies.

Smaller detection capability generally permits finer protection, but it may affect operating range, installation tolerance, beam count, price, and resistance to contamination.

Protective height

Protective height is the vertical span covered by the active sensing field. It must match the opening that needs protection—not merely the machine’s overall height.

A 900 mm curtain does not protect a 1,200 mm opening unless additional fixed or electro-sensitive safeguards cover the remaining path.

Type 2 versus Type 4

IEC 61496-1:2020 sets general design, construction, and testing requirements for non-contact electro-sensitive protective equipment. IEC 61496-2:2020 adds requirements for active opto-electronic protective devices used to detect people.

Type classification concerns the protective device’s safety-related behavior and fault-detection requirements. It is not another name for beam spacing.

The detailed comparison in Type 2 versus Type 4 safety light curtains is useful when a risk assessment must distinguish lower-integrity applications from machines where severe or irreversible injury is credible.

My view is straightforward: when crushing, amputation, or permanent disability is a realistic outcome, procurement should not begin by asking which type is cheaper.

It should begin by establishing the required safety function.

What the Enforcement Files Reveal

The accident record exposes the gap between owning a light curtain and operating a protected machine.

A first-day worker lost three fingers

In November 2022, a new employee at United Hospital Supply Corp. suffered the amputation of three fingers while operating a press brake. OSHA said supervisors and employees had deliberately bypassed the light curtain.

The May 17, 2023 enforcement release reported three willful violations, 17 serious violations, one other-than-serious violation, proposed penalties of $498,464, and placement in OSHA’s Severe Violator Enforcement Program.

The curtain was present.

The protection was not.

A supervisor disabled a light curtain

In 2019, OSHA investigated a New Hampshire furniture manufacturer after an employee was pulled into an automated woodcutting machine and seriously injured.

Investigators found that a supervisor had disabled the light curtain, preventing the machine from stopping when a person approached the point of operation. OSHA cited one willful and 36 serious violations, with proposed penalties totaling $378,488.

This was not a hidden electronic defect.

It was a management decision.

A disabled curtain preceded a partial finger amputation

An April 22, 2016 incident at a Wisconsin bag manufacturer left a worker with a partially amputated right index finger while clearing a jam from a bag-sealing machine. OSHA found that manufacturer-installed light curtains had been disabled.

Clearing jams is exactly when workers enter spaces that production planning pretends they will never enter.

That is why risk assessments based only on normal automatic operation are weak. Cleaning, adjustment, troubleshooting, setup, recovery, tool changes, and maintenance must be considered.

The numbers are not abstract

The U.S. Bureau of Labor Statistics recorded 6,200 work-related amputations involving days away from work in 2018. Machinery was involved in 58%, or 3,580 cases, and the median recovery period was 31 days, compared with nine days across all injury types.

Metalworking, woodworking, and special-material machinery accounted for 1,660 of those amputation cases.

More recent BLS reporting shows private-industry employers recorded approximately 2.5 million nonfatal workplace injuries and illnesses in 2024. That broad number does not isolate machine-guarding failures, but it is a useful reminder that lower aggregate injury rates have not made bypassed safeguards acceptable.

Where Machine Guarding Light Curtains Work Best

Safety light curtains are especially useful where workers need frequent access and a fixed barrier would interfere with loading, unloading, inspection, or normal production.

Common applications include:

• Mechanical and hydraulic presses
• Press brakes
• Stamping cells
• Robot loading areas
• Packaging machinery
• Assembly stations
• Palletizing systems
• Automated cutting equipment
• Injection-molding access points
• Conveyor transfer areas
• Winding and converting machinery

For large presses, vibration, exposed structures, long stopping times, coolant, oil mist, and physical impact can change the product requirement. A standard slim housing may work electrically and still be a poor mechanical choice.

That is where heavy-machine safety light curtains deserve separate consideration.

Machines with front, side, and rear access may require multiple coordinated sensing fields rather than a single curtain across the most obvious opening. The site’s multi-sided access protection systems address this broader geometry.

And sometimes a light curtain is simply the wrong safeguard.

Use physical guards, interlocked doors, trapped-key systems, pressure-sensitive devices, scanners, or combinations of technologies when the hazard involves:

• Flying material
• Broken tooling
• Hot fluids
• Sparks or radiation
• Long stopping times
• Fall-through or climb-over access
• A person remaining undetected inside the guarded area
• A machine that cannot reliably stop mid-cycle

How to Select the Best Safety Light Curtain for Machine Guarding

The best safety light curtain is not the model with the highest beam count, the smallest enclosure, or the most impressive Type marking.

It is the model that fits a validated safety function.

Before requesting a quotation, document at least the following:

1. Machine type and operating cycle
2. Hazardous motion and injury severity
3. Required Performance Level or SIL target
4. Measured worst-case stopping time
5. Intended finger, hand, arm, or body detection
6. Required protective height
7. Distance between emitter and receiver
8. Available mounting space
9. All possible access routes
10. Output type and safety-controller compatibility
11. Reset method and reset-button location
12. Need for EDM, muting, or blanking
13. Exposure to dust, oil, water, vibration, welding sparks, or reflective material
14. Destination-country standards and machine-specific requirements

Do not send a supplier only the opening width and ask for “a suitable curtain.”

That is not enough information to select a personnel-protection device responsibly.

FAQs

How do safety light curtains work?

Safety light curtains are electro-sensitive protective devices that use an emitter and receiver to create a field of infrared beams; when a person interrupts one or more beams, the device changes its redundant safety outputs so the machine’s safety-related control system can stop or prevent hazardous motion.

The stop command normally travels through dual OSSD channels to a safety relay or safety PLC. That controller then operates contactors, monitored valves, a clutch-brake system, or a drive safety function such as Safe Torque Off.

Do safety light curtains stop a machine instantly?

Safety light curtains do not stop a machine instantly; they detect an intrusion within their specified response time and initiate a stop command, while the machine’s controller, switching devices, brake, drive, hydraulic system, or pneumatic system still need additional time to bring hazardous motion to a safe state.

The mounting distance must account for the entire response chain. A 10- or 15-millisecond sensor can still be unsafe when installed too close to machinery that needs several hundred milliseconds to stop.

What is the difference between a safety light curtain and a standard light curtain sensor?

A safety light curtain is designed and tested as part of a safety-related personnel-protection system with monitored outputs, fault-response behavior, and defined integration requirements, while a standard light curtain or optical grid is normally intended to detect, count, measure, position, or inspect products within an automation process.

A standard sensor may look nearly identical from outside. Appearance is irrelevant. Its safety classification, diagnostics, outputs, documentation, and intended use determine whether it belongs in a personnel-protection circuit.

How far should a safety light curtain be from the machine hazard?

A safety light curtain must be positioned far enough from the hazard that the complete machine reaches a safe state before a person can travel from the sensing field to the danger point, using the applicable safety-distance standard, measured stopping time, approach speed, sensor response, control delays, and penetration allowance.

ISO 13855:2024 is the current international positioning standard. In the United States, machine-specific OSHA and ANSI requirements may also apply. The distance should be calculated and validated for the actual machine rather than copied from another installation.

Can a machine restart automatically after the light curtain clears?

A machine should not automatically restart merely because the safety light curtain becomes clear when a person could have entered or remained in the safeguarded area; restart behavior must follow the risk assessment, applicable standards, operating mode, visibility of the protected space, and validated manual-reset or presence-detection strategy.

A reset button should not be reachable from inside the danger zone, and the person resetting the system should be able to inspect the safeguarded space. Larger cells may need additional inside-area detection or trapped-person controls.

Is a Type 4 light curtain automatically a PL e safety system?

A Type 4 light curtain is a high-integrity active opto-electronic protective device classification, but it does not automatically make the complete machine safety function PL e or SIL 3 because the achieved performance also depends on the controller, wiring, diagnostics, output devices, actuators, software, stopping performance, fault handling, and validation.

A premium sensor connected to an ordinary PLC and one unmonitored contactor is not a high-performance safety architecture. The claim belongs to the complete validated safety function, not one component.

Turn the Machine Data Into a Defensible Safety Specification

Do not choose a safety light curtain from beam count and price alone.

Measure the machine’s worst-case stopping time. Identify every access route. Define the body part that must be detected. Establish the required PL or SIL. Document the environment, mounting distance, protective height, reset method, output architecture, and any muting or blanking requirement.

Then select the device.

For an application review, product recommendation, OEM project, retrofit, or technical quotation, send those details through the Safety Curtain engineering contact page. The more complete the machine data, the more defensible the final selection will be.