May 27, 2026 · Safeguarding
Every fenceless or partially fenced robot cell in a West Michigan plant relies on presence-sensing safeguards. The same is true for many fully fenced cells with an operator load station, a material entry conveyor, or a maintenance access aperture cut into the perimeter. Light curtains and laser scanners are the two devices that do the work, and their correct selection, placement, and integration are the single most common gap we find on robotics safety audits across Grand Rapids, Holland, Kalamazoo, and Battle Creek.
This guide walks through how a senior controls and EHS engineer should think about these devices. Selection criteria, the math behind safety distance, the difference between muting and blanking, the control circuit category required, and the integration mistakes that pass installation but fail a real audit.
The two devices are not interchangeable, and the right choice depends on the geometry of the access being protected, not on cost or familiarity.
A light curtain projects a vertical (or horizontal) plane of infrared beams between a transmitter and receiver. Anything that breaks any beam in the plane triggers an output. The protected area is, in practical terms, a flat curtain at the plane of the device. That makes light curtains the right answer for a defined opening: an operator load station, a conveyor entry into a cell, an access aperture cut into a fenced perimeter for tool changes or fixture loading.
A laser scanner emits a rotating laser pulse and measures time-of-flight to anything in its field of view across a horizontal sweep, typically 190 to 275 degrees. The protected area is a two-dimensional zone on the floor, configurable into shaped warning and stop fields. That makes scanners the right answer for area protection: the floor space around a fenceless cell, the maintenance corridor at the back of a cell, the operator approach zone for a collaborative or semi-collaborative installation.
Many real cells use both. A SICK or Banner light curtain at the operator load station handles the opening. A SICK microScan3 or Keyence SZ-V laser scanner protects the rear floor area where a maintenance technician approaches from a less defined direction. The two devices feed the same safety controller, with field configurations that match the muting requirements and the production sequence.
Resolution defines the smallest object the device will reliably detect. Choosing the wrong resolution is one of the most expensive mistakes in safeguarding, because the device either over-builds for the application or fails to detect the relevant body part.
| Resolution | Detection target | Typical use |
|---|---|---|
| 14 mm | Finger | Press brakes, point of operation, small parts handling within reach of the hazard |
| 30 mm | Hand | Most robot cell access openings, manual load stations, assembly cells |
| 40 to 70 mm | Arm or leg | Body access points where reach-through is not a concern |
| 70 to 90 mm (body) | Whole body | Perimeter scanners, large fenceless cell zones, walk-in cells |
For most West Michigan robot cells, hand detection at 30 mm is the working default for openings within reach of the robot envelope. Anywhere a worker could reach in and contact moving equipment, 30 mm is the floor. The temptation to specify a coarser body-detection curtain because the access is large enough to walk through is the trap; ANSI/RIA R15.06 ties resolution to the part of the body that must be detected before the hazard, not the geometry of the opening alone.
A light curtain mounted right at the hazard would let an intruder reach the hazard before the cell stopped. The safety distance is the minimum standoff between the curtain plane and the nearest hazard, calculated so that the cell completes its stop before reach is possible.
ISO 13855 gives the formula:
The variables:
A realistic example. A robot cell with a stop time of 350 ms (robot 250 ms, safety relay 30 ms, curtain response 70 ms), protected by a 30 mm resolution curtain mounted vertically, with hand approach at 1600 mm/s, calculates:
That is the minimum distance from the curtain to the nearest part of the robot envelope. Most installations get this number wrong on the first pass, usually by under-counting the cell stop time, ignoring the response time of the safety controller, or skipping the intrusion factor. The fix on a real cell is often as simple as moving the curtain back 200 mm to satisfy the calculation, but it has to be calculated first, and the calculation has to be documented.
A laser scanner gives the cell something a light curtain cannot: a graduated response. The same scanner can monitor multiple zones simultaneously and output different signals for each. The standard configuration is a two-field approach: a warning zone and a stop zone.
The warning zone is the outer area. When breached, the scanner triggers a controlled slowdown of the robot, a visual or audible alert, and often a hold on automatic restart logic. The worker is not yet at the hazard. The cell has time to react in a measured way.
The stop zone is the inner area, calculated under ISO 13855 with the same formula above. When breached, the scanner triggers a Category 0 or Category 1 stop per ISO 13849-1 and ISO 10218, depending on the cell architecture. The worker is now close enough that the cell must come to a stop before reach.
This graduated response reduces nuisance stops. A fenceless robotic palletizer cell sees a forklift driver pass within five feet several times per shift. A single-zone scanner stops the cell on every pass. A two-field scanner slows the cell into the warning zone and only stops it if the driver actually approaches the hazard. The production benefit is real, and it is achievable without compromising safety, but it requires the safety controller to handle the graduated signals correctly.
Both light curtains and scanners need a way to allow expected material flow without stopping the cell every time a pallet of product breaks the field. The two techniques are muting and blanking, and they are not the same.
Muting uses dedicated muting sensors (typically two pairs of photoelectric sensors or a single retroreflective array) that confirm the object breaking the curtain is the expected material in the expected direction, not a person. The muting circuit temporarily disables the curtain for that pass, then re-enables it. Muting is dynamic and conditional. It is the right answer for an outfeed pallet conveyor.
Blanking permanently ignores a defined portion of the protected field. Fixed blanking is used when a piece of equipment (a conveyor frame, a fixture, a structural member) permanently sits inside the protected area. Floating blanking allows a small variable region within the field, useful when the expected material varies in position slightly. Blanking should be configured by trained personnel, documented, and locked in the safety controller.
The two failure modes we find most often: muting sensors mounted in the wrong sequence or with timing windows wide enough to allow a person to walk through during a muted window; and blanking configured loosely, with too much of the protected field permanently disabled, so the effective protection is smaller than the cell drawing suggests. Both pass a casual install check. Both fail a real safety audit.
Picking the right curtain or scanner is the easy half. Integrating it into a control system that meets the required Performance Level under ISO 13849-1 (or SIL under IEC 62061) is where most West Michigan robot cells fall short.
For a typical industrial robot cell with hazardous motion, the risk assessment usually drives the safety function to Category 3 with Performance Level d (PLd), or Category 4 with Performance Level e (PLe). That requires:
A light curtain wired into a standard relay, with no safety controller, no cross-monitoring, no diagnostic coverage, is the most common audit finding we write up. The device itself is correct. The integration does not meet the category. The fix is rarely a hardware swap of the curtain; it is the installation of a safety controller and proper dual-channel wiring. Treating the safeguard as a sensor instead of a safety function is the conceptual error behind most of these gaps. The full alignment between ANSI/RIA R15.06, ISO 10218, and OSHA 1910.147 is covered in our standards alignment guide.
Each of these is a citable condition under OSHA 1910.147, MIOSHA Part 85, and the consensus standards. Our robotics safety gap analysis closes the most common combinations in a single site visit, with a documented test of safety distance and a written gap report.
The robot cells we audit across West Michigan auto suppliers, food processing, plastics, and packaging plants pull from a short list of devices:
The brand matters less than the integration. A correctly integrated entry-tier device outperforms a top-tier device wired into a non-safety relay. The integrator and the documentation are what determine whether the cell actually meets the required category. Full technical references on these standards are available at the OSHA robotics safety page and through the ISO 10218-1:2025 standard listing.
We will walk your robot cell, verify the safeguarding hardware, recalculate safety distance under ISO 13855, validate the muting and blanking configuration, and confirm the control architecture meets the required PL or SIL. Direct line: 616-217-3325.
Request Gap AssessmentLight curtains protect a defined opening, like an access aperture in a guarding fence, where the geometry of the entry is fixed. Laser scanners protect an area in two dimensions, like the floor space around a fenceless or partially fenced cell, where personnel can approach from multiple angles. Many West Michigan robot cells use both: a curtain at the operator load station and a scanner monitoring the rear of the cell where service access happens.
It depends on what the curtain is protecting. Finger detection requires 14 mm resolution. Hand detection requires 30 mm. Body or torso detection on a perimeter installation allows up to 70 to 90 mm. ANSI/RIA R15.06 and ISO 13855 tie resolution to safety distance: tighter resolution means the curtain can be mounted closer to the hazard. The wrong resolution either over-builds the cost or under-protects the access point.
ISO 13855 gives the formula: S equals K multiplied by T plus C. K is the human approach speed, typically 1.6 or 2.0 meters per second. T is the total stopping time of the cell, including curtain response time, control system response, and machine stop. C is an intrusion factor based on the curtain resolution. A typical industrial robot cell calculates a safety distance of 300 to 800 mm depending on stop time and resolution.
Both temporarily allow material or a known object to pass through the protected field without stopping the cell, but they work differently. Muting uses external sensors that confirm a pallet or product is what is breaking the beam, and disables the curtain only for that pass. Blanking permanently ignores a defined portion of the field, used when fixed equipment or a conveyor permanently sits inside the protected area. Muting must be control-reliable. Blanking must be configured by trained personnel only.
No. Presence-sensing safeguards protect operators during normal production operation by stopping motion when the protected field is breached. They are required by ANSI/RIA R15.06 and ISO 10218 for many robot cells, and they are valuable. They are not energy isolation. OSHA 1910.147 service and maintenance work requires actual lockout at a disconnect. Light curtains protect the operator running the cell; LOTO protects the worker inside the cell during service.
ISO 13849-1 typically requires Category 3 or Category 4 architecture with Performance Level d or e for robot cell safeguarding, depending on the risk assessment. Equivalent IEC 62061 ratings are SIL 2 or SIL 3. This means dual-channel monitoring, cross-checking, and fault detection on the safety circuit. A light curtain wired into a basic relay without a safety controller fails this requirement, and is one of the most common integration gaps we find on audits.
Related reading: Robot Cell LOTO: Aligning ANSI R15.06, ISO 10218, and OSHA 1910.147, Collaborative Robot Safety per ISO/TS 15066, Robot Cell LOTO Procedures.