Warehouse Robot Safety Laser Scanner & LiDAR Sensor Matrix: The Complete Pairing Guide

Written by the QJKH engineering team · Reviewed by CCH Shanghai Sensing Intelligence safety sensing engineers · 2026

Warehouse robot safety laser scanner choices decide if a cell passes its IEC 61496 audit or remains grounded in commissioning hell.Most integrators we meet still buy “a sensor” when the job requires two vastly different devices with different certifications, different optics, and different wiring – one to guard humans, one to guide motion. This guide is the reference table we wish we had had when we started matching sensors to warehouse robots ten years ago.

It absorbs palletizing, picking and sorting, loading and unloading robots, stacker cranes, pallet shuttles, case shuttles, overhead hoists, automated case-handling mobile robots (ACRs), and rail-guided vehicles (RGVs, and its automated guided vehicle cousin the AGV) into a single pairing sheet, then walks through the standards, environmental constraints, and six questions we use to arrive at an ironclad choice in less than ten minutes.

Why Warehouse Robot Safety Sensing Splits Into Two Tracks

Every warehouse robot — from stationary and mobile palletizing cells to autonomous mobile robots (AMRs) roaming open aisles — holds two distinct sensing roles, and one sensor simply cannot fulfill both duties efficiently. Safety laser scanners (also called area scanners in older literature) are designed to safeguard humans and prevent dangerous motion with two programmable safety zones — a protection zone (the hard OSSD trip) and a warning zone that reduces motion velocity before intrusion, with the protection zone sometimes subdivided into dual zones for complex cell geometry. Navigation LiDAR, by contrast, is designed to construct a point cloud, feed fleets the position data they need, and avoid static non-human obstacles at high speed along the robot’s direction of travel — roles a safety scanner cannot do because its response time and field-of-view are optimized for maximum protection, not mapping.

This confusion is not just academic. Both the OSHA robotics overview y el OSHA Technical Manual Section IV, Chapter 4 impose safeguarding responsibility on the machine builder and robot application integrator. If you specify a navigation LiDAR because a datasheet somewhere called it a “laser scanner,” no OSHA inspector will accept it in a PLd audit.

Global deployment numbers make the stakes concrete. According to the IFR World Robotics 2024 report, 4,281,585 industrial robots are now operating in factories worldwide — a 10% year-over-year increase, with annual installations exceeding half a million units for the third consecutive year. A meaningful share of those deployments are now mobile or semi-mobile warehouse platforms, which means every warehouse robot safety laser scanner specification error compounds across thousands of new cells every year.

The Warehouse Robot Sensing Pairing Matrix

This is the core of this guide. Instead of define sensors abstractly, we pairing each warehouse robot type with the sensor class and specific QJKH model that aligns with its threat profile, its motion envelope, and its environment. The matrix uses the QJKH SH27 safety laser scanner (IEC 61496 Type 3, SIL 2, PL d, Cat. 3) series for all fixed safety functions and the QJKH YB27 navigation LiDAR for all position, mapping, and obstacle-avoidance functions. As each cell varies, treat the model column as a baseline rather than a definitive prescription.

What type of sensor does an AGV actually need?

An AGV needs both. The safety laser scanner takes care of Type 3 protective-field tasks at the front of the vehicle to prevent the vehicle from striking anyone and the navigation LiDAR manages path-following, obstacle-avoidance around pallets and racks, and position feedback to the fleet manager. As the sensors run concurrently on separate harnesses, the scanner’s OSSD output wires into the safety PLC or safety relay, and the LiDAR’s Ethernet UDP stream feeds the motion controller. Configuring just one device to accomplish both jobs only works on very slow, fenced-cell AGVs—and even then, the stationary-cell safety laser scanner still handles cell entry because a navigation LiDAR cannot validate down to PL d.

A few notes before making use of the table. First, the SKUs quoted expect typical warehouse conditions (indoor, 3,000 lux ambient at the scanner window, 1.0-1.5 m/s maximum vehicle speed, 5-45 C.) If those parameters are exceeded, either upgrade the protective radius or move onto the dock-door installation pattern we specify in the environmental section. And second, the matrix considers the picking and sorting robot safety laser scanner and the robotic sorter safety laser scanner to be the same class, as both have operator hand-over zones and both are permanently fitted to the cell. The actual distinction is in the zone library programming, not the hardware.

Everything below that – the standards section, the dual-stack wiring discussion, the decision tree – exists to substantiate and improve upon the decision matrix presented here. For more information on navigation LiDAR specifications and range, see our guide to navigation LiDAR range and resolution specs for AGV and AMR platforms which covers the YB27 matrix row by row.

Functional Safety Standards (Why a Nav LiDAR Cannot Replace a Safety Scanner)

Safety laser scanners are not defined by their optics. They are defined by the four certifications stamped on the datasheet. If any one is missing, you do not have a safety scanner – you have a measurement LiDAR with marketing aspirations. Our QJKH SH27 series publishes the full stack, and that stack is the industry baseline you should demand from every vendor you investigate.

IEC 61496-1 / -3: Type 3. Type 3 is the laser safety baseline for electro-sensitive protective equipment (ESPE) — the same standard family that covers light curtains and pressure mats. It indicates the scanner has been tested for fault detection, signal diversity, and electromagnetic compatibility, with warning fields and protection fields both validated against object-resolution targets. Type 3 is the appropriate class for warehouse robot cells; Type 4 is for more hazardous fixed-guarding applications, such as press-brake light curtains, where intrusion into hazardous areas carries higher consequence.

IEC 61508: SIL 2. Safety Integrity Level 2 indicates the probability of dangerous failure per hour (PFH) lives in the 10⁻⁷ to 10⁻⁶ range — a statistically bounded reliability goal linked to a quantified hazard analysis. No SIL rating is published for navigation LiDAR because its failure modes are not analyzed against a safety function.

ISO 13849-1: PL d, Cat. 3. Performance Level d with Category 3 specifies the architectural requirements — dual-channel signal paths, diagnostic coverage, and the ability to detect common-cause failures. Your integrator will mention this standard when performing the cell’s risk assessment. Third-party datasheets from multiple safety sensor manufacturers display the same PL d / Cat. 3 rating, confirming this is the industry baseline — not a unique value.

ISO 10218-2. This is the mandate for industrial robot system safety. Here is where the cell design — which includes the selection of presence-sensing protective equipment — is rigorously inspected. Any scanner meeting IEC 61496-3 Type 3 qualifies as complying with the PSPE requirement under ISO 10218-2, assuming proper mounting geometry and stop-chain timing. For a full breakdown of these four standards as they apply to scanner selection, see our deep dive on the industrial safety laser scanner standards and selection criteria.

Navigation LiDAR for Stacker Cranes, Shuttles, and Overhead Hoists

In the case of rail-guided and overhead-rail platforms, the safety requirement curve shifts. Because vehicle motion on these platforms is limited to a linear aisle, the human collision potential is lower than for a free-roaming AMR. These platforms require high-accuracy position feedback — fraction-of-a-degree angular resolution and stable range at long distances — which is well provided by a direct time-of-flight (dToF) single-line navigation LiDAR.

Stacker crane LiDAR sensors are typically long-range, high-resolution variants mounted on the carriage, supporting two functions: absolute position along the aisle (for put-away and retrieval targeting) and end-of-aisle gap measurement. Our QJKH YB27-40HE (40 m measurement range, 0.2° resolution at 50 Hz, Ethernet UDP output) is a common choice because 40 m comfortably covers the longest practical aisles, and the Ethernet output lets the WMS read position directly.

Pallet shuttles and case shuttles inhabit even tighter geometry. Pallet shuttle LiDAR sensors face two constraints: mounting height generally stays below 150 mm over the rails, and the sensor must see the far end of a 15–25 m tunnel with pallets stacked above it. YB27-15CS (15 m, 0.3° resolution at 30 Hz) solves the short-shuttle case, and YB27-25HS handles the longer racking. Case shuttles drive the data rate higher as pick-to-place cycles are faster — our dual-output YB27-15CD (Ethernet + PNP) can both order the controller to track position and trip the emergency stop relay on a fast obstacle.

Overhead hoist LiDAR is another dimension of the problem. OHT vehicles run alongside ceiling-guided rails in semiconductor fabs and bonded warehouses. They serve as vertical gap measurement and wafer-FOUP clearance sensors — not as people-avoidance sensors, since the rail keeps the vehicle above human height. Our YB27-25HE handles the variable fluorescent lighting these environments impose, with 100,000 lux immunity, and 25 m range matches typical inter-rail segment lengths.

“The greatest point of failure we find on stacker cranes is not in the scanner but in the singular decision to use one sensor for both navigation and emergency stop. When a technician has to perform maintenance at the end of the aisle, the nav LiDAR cannot verify the gap is safe – only a Type 3 safety scanner mounted at the end-stop can.”

RGV and ACR Sensor Pairing — The Dual-Stack Case

RGVs and ACRs are the clearest demonstration of why one sensor cannot cover both jobs on a single platform. RGVs move linearly but can travel 2–3 m/s along a corridor where maintenance staff periodically need access. ACRs roam freely through aisles where human pickers work on the same rack face. In both cases, navigation LiDAR cannot satisfy ISO 13849-1 PL d, and a safety laser scanner cannot deliver the 0.1° angular resolution required for natural-feature SLAM.

The clean architecture for these platforms yields a dual stack. Our RGV lidar – typically a YB27-40HE at the front and back – gives a point cloud over Ethernet UDP to the motion controller. Our safety laser scanner for RGV – the SH27-05D at each front and rear end – drives two OSSD outputs into a safety PLC or dedicated safety relay module. These two stacks do not compete with each other. Navigation operates in the soft real-time loop, safety operates in the hard real-time loop, and the stop decision takes place solely within the safety PLC.

How do you wire a safety laser scanner and a navigation LiDAR on the same mobile robot?

The 17-pin M12 power connector on our QJKH SH27-03D scanner supplies DC 24 V, two OSSD channels (PNP, 24 V high when clear), and EDM feedback from the downstream contactors. Those OSSD lines feed the safety PLC input module — not the motion controller. A separate M12 4-pin Ethernet connector supplies the diagnostic stream over TCP/IP. Our YB27-25HD navigation LiDAR outputs its Ethernet UDP stream to the motion controller and, in dual-output mode, also feeds four configurable PNP outputs into the safety relay as a secondary slow-down source. Both devices sit on a fused distribution block sharing a 24 V rail, but nothing else, to preserve the Category 3 dual-channel architecture. See our reference on safety relay modules for OSSD wiring on mobile robots for hookup details.

Environmental and Installation Constraints You Cannot Ignore

Datasheet specs are just that, if the scanner works in the lay out you want, it will work when you put it there. But five environmental conditions will be factors in every real deployment, and determine if your purchase is still a good one at commissioning.

Ambient light. Safety laser scanners publish an ambient-light immunity of roughly 3,000 lux. Navigation LiDAR tops out much higher — 100,000 lux on our QJKH YB27 series. This is not marketing speak. Installation manuals from Rockwell, Banner, SICK, and Omron all explicitly warn against mounting safety laser scanners where incident sunlight can saturate the window, because saturation produces false trips, and a false trip at a dock door means the fleet sits idle until somebody reaches for the reset. If your loading and unloading robot safety scanner has to look at an open dock door between 10 a.m. and 2 p.m., either add a sun shield, reorient the scanner, or use a navigation LiDAR for the approach segment and bring the safety scanner online only when the door is closed.

Temperature and condensation. SH27 operating temperature sits at –10 to +50 °C with no frost or condensation. Refrigerated warehouse deployments below –10 °C fall outside this envelope and need either a heated enclosure or a different sensor class entirely.

IP rating. Both SH27 and YB27 publish IP65 as the baseline. That means dust-tight and protected against water jets, but it is not rated for high-pressure washdown or submersion. Spray-down food-grade environments need additional protection.

Vibration and impact. SH27 vibration class sits at 10–55 Hz at 0.35 ± 0.05 mm, 5M1 per IEC 60721-3-5. Shock test is 10 g over 16 ms. Mounting a safety scanner on a forklift attachment or on a stacker crane carriage — where you will see frequent 3–5 g starts and stops — works within spec, but the tolerance zone and ZR extension (each 350 mm on the SH27) were implemented because the field under vibration is less stable than at rest.

Mounting geometry. Protective fields assume a specific mounting height and angle, and blind spots appear wherever the scan plane misses an approach vector. At 70 mm minimum detectable object resolution across the maximum protective radius, you need an ankle-height scan plane to mark the boundary between walkable floor and danger zones; a mid-leg scan plane misses children and stooped operators. One more muting detail worth naming: when a conveyor needs to cross the protective field (a loading and unloading robot dropping cases onto an outbound line), use the muting function (configured to mute the protection zone only when the conveyor’s paired safety sensors agree) so the scanner ignores the pallet but not the person walking next to it. Commissioning checklists should verify the actual scan plane, not assume the datasheet value.

Selection Decision Tree — Six Questions in Ten Minutes

Use these six questions in order. Whichever one you answer “yes” to first determines the sensor class; later questions draw in the specific model. This is the process our QJKH engineering team guides new integrators through and it reliably returns a defensible choice faster than arguing over datasheets.

1. Is the robot rail-guided or free-roaming? Rail-guided (stacker crane, shuttle, OHT) favors navigation-LiDAR-dominant; free-roaming (AMR, ACR, forklift) favors safety-scanner-dominant. RGVs occupy the middle ground and get the dual stack.
2. Does the cell or route need SIL 2 / PL d certification? If yes — and for any application where a human may enter the motion envelope it is yes — a safety laser scanner is required. A navigation LiDAR cannot replace it.
3. What protective radius is required? Calculate (vehicle speed × stop chain time) + safety distance according to ISO 13855. Under 3 m → SH27-03D. 3–5 m → SH27-05D. Over 5 m → you must slow, or add a second scanner.
4. Is the installation exposed to above 10,000 lux ambient light? If yes, the approach segment is a navigation-LiDAR task (YB27, 100,000 lux) and the safety scanner only engages once the vehicle reaches the end of the direct light.
5. Does the fleet manager require immediate position data from this platform? If yes, select a YB27 dual-output variant (CD/HD) so one device supplies both Ethernet UDP positioning and PNP obstacle trips.
6. Is the output path managed by a safety PLC or a dedicated safety relay? Either path functions. Our SH27 OSSD lines terminate on a 17-pin M12 connector and connect directly into any modern safety input module. If designing a standalone cell, our QJKH SH27 safety laser scanner series pairs natively with the corresponding safety relay modules.

A brief note for the pallet-shuttle and case-shuttle car readers who jumped to this section: the optimal warehouse robot safety laser scanner for narrow-aisle pallet shuttles is almost always the SH27-03S since 3 m protective fields match the aisle width, the compact housing fits under 150 mm clearance and the OSSD output wires into the shuttle’s on-board safety PLC without a dedicated controller. Navigation duty on the same shuttle is assigned to the YB27-15CS.

Preguntas frecuentes

What is the difference between a safety laser scanner and a navigation LiDAR?

Ver respuesta

A safety laser scanner is an IEC 61496 certified (Type 3, SIL 2, PL d) product: it trips an OSSD output when persons or objects intrude into a controlled field,with the intended purpose of stopping motion before impact. The navigation LiDAR is a measuring instrument: high-resolution, longer-range, non-safety IEC 61496 certified, with applications in SLAM, path following, and benign obstacle avoidance. You cannot mix and match these functions in a safety application and most heavy-duty warehouse robots require both.

Can one sensor cover both safety and navigation on an AMR?

Ver respuesta

No. Both tasks have incompatible IEC 61496 certification and optical tuning requirements. When marketing dual-out, multi-channel LiDAR modules as combined solutions, the safety channel is usually run at lower resolution than the navigation channel, and neither channel has a single, integrated IEC 61496 certification. Specify individually-certified devices.

What protective radius do I need for a palletizing cell?

Ver respuesta

Apply the ISO 13855 safety distance formula — (speed of approach × total response time) + intrusion distance — and round up. Most fenced palletizing cells land at 2.5–4 m, which is why the SH27-05D (5 m protective) covers the common case with a comfortable margin.

Are safety laser scanners required by law for warehouse robots?

Ver respuesta

The short answer is no, but OSHA and relevant machine-directive regimes (EU Machinery Directive, ANSI/RIA R15.06) expect the designer/commissioner to conduct a risk assessment to determine what kind of safeguarding is tailored best to the device, the location, and the use. For any cell where a person can enter the motion envelope, a Type 3 presence-sensing device is almost always what the integrator recommends, and a safety laser scanner is still usually the easiest. The various legal liabilities involved are on the system integrator.

How often do safety laser scanners need recertification?

Ver respuesta

The document-supplied certificate has a non-expiring date, but the cell-system-level risk analysis should be reviewed whenever the layout, travel speed, or robot work program changes. Window-cleaning and contamination indicators should be checked on every shift.

Why is 3,000 lux the light-immunity cutoff on safety scanners?

Ver respuesta

3,000 lux is not a physics ceiling – it is the illuminance level at which the device can still be licensed to meet the fault-detection and false-trip criteria accepted under IEC 61496 by Type 3 equipment. Above that there are technical interactions between the receiver’s signal-to-noise ratio and the level at which a false trip is guaranteed; below it are physical interactions between the sender’s optics and the receiver’s far-field inability to distinguish objects and noise, so the scanner disables the LIDAR’s fault-triggering ability and effectively switches to a safe-state OSSD-Off. A safe navigation LIDAR operating at 100,000 lux immunity does not encounter this restriction because it is working outside of a safety application.