In short: Safety laser scanner faults fall into four buckets — optical, mechanical, electrical, and logical/configuration. False trips are almost always optical (contamination, weather, low-reflectivity objects, scanner cross-talk), and they self-clear. A lockout is different: it is a latched safe-state from an OSSD self-test, EDM mismatch or hardware diagnostic, and it must never be reset without understanding the cause. Clean the window on a schedule derived from the contamination warning, not the calendar. Treat OSSD, EDM and CRC errors as audit events — log them and root-cause them before you press reset.
A maintenance lead at a logistics centre put it to us neatly last month: “The scanner trips, we press reset, the line runs. The next shift the scanner trips, they press reset. Nobody knows whether anything is wrong or not.” That is the exact failure pattern this article is written to break. A safety laser scanner has good diagnostics; the diagnostic is only useful if someone reads it.
This is a working reference for engineers and maintenance teams who are responsible for industrial safety LiDAR scanners installed under IEC 61496-3 with safety functions evaluated to ISO 13849-1. It assumes the scanner is already commissioned and the safety function is documented; what we are dealing with is the year-three reality of dust, vibration, contactor wear and the slow drift that nobody wrote down. The pillar reference across the LiDAR family is our Industrial Safety LiDAR — Complete Reference; if you have not been there, start there and come back.
A failure-mode taxonomy that actually helps
We learnt years ago that the fastest diagnosis comes from deciding which of four buckets the fault is in before opening any wiring panel. Skip this step and you will spend an afternoon checking cables on what turns out to be a dirty window.
| Bucket | Examples | How the scanner tells you |
|---|---|---|
| Optical | Window contamination, fog/dust, ambient IR, retro-reflective surfaces, low-reflectivity targets | Contamination warning, intensity/quality diagnostic, intermittent OSSD |
| Mechanical | Mounting drift, vibration, fixture loosening, impact damage, thermal cycling | Field-boundary drift, alignment warning, sudden onset trips |
| Electrical | OSSD short-circuit, EDM feedback fault, ground loop, broken sync wire, supply ripple | Lockout codes, OSSD test pulse failure, controller event log |
| Logical / configuration | Wrong field set, EDM enabled with no feedback wired, checksum mismatch, response-time mis-set | Lockout on power-up, version/CRC error, refused parameter download |
The taxonomy maps loosely onto the IEC 61496-3 fault model: a Type 3 scanner is expected to detect and respond to its own single faults inside the published fault response time, and any fault it cannot detect must be tolerable in light of the architecture. What that means in practice: the scanner is doing a great deal of self-checking, and most of what you see as a fault is the scanner correctly refusing to keep running.
False trips — the full taxonomy
A “false trip” is a misnomer; the trip is real, but the object that caused it is not what you wanted to protect against. Sorting these out is mostly a question of pattern recognition.
| Trip source | Mechanism | Standard fix |
|---|---|---|
| Precipitation (rain, snow, fog) | Near-range backscatter creates spurious returns | Verify multi-echo enabled; check window cleanliness; consider hood/shroud |
| Airborne dust / fines | Same as fog; worse with surface buildup | Clean window; shorten cleaning interval; add positive-pressure purge if available |
| Low-reflectivity object | Black rubber, dark fabric returns too little energy near field edge | Use scanner with 1.8% remission spec; size the field with margin from the rated range |
| Ambient IR (sunlight, halogen, arc) | Receiver saturation by broadband light | Shroud receiver from direct sun line; relocate or shield arc sources |
| Scanner-to-scanner cross-talk | Adjacent scanner's pulse mistaken for an echo | Use modulation/coding; offset scan planes; hardware sync where supported |
| Retro-reflective / mirror surfaces | Multipath returns from polished floors, glass, mirrors | Matte coating, baffles, rotate scanner; configure reflector blanking only when certified |
| Mounting vibration | Field shifts physically into a static obstacle | Stiffen bracket, isolate from machine shock, re-teach field with machine running |
A note on multi-echo. Modern safety scanners receive several reflections from each emitted pulse and decide which one represents the object. The first echo will often be fog, dust or rain near the housing; the algorithm prefers the last echo, which is the solid target behind. Multi-echo is the single biggest reason a modern scanner survives a workshop full of welding smoke or a coal-yard breeze where older units would have nuisance-tripped. Two things you should still check: that the feature is enabled in the parameter set, and that the front window is clean enough that the near-range echoes have not started overwhelming the receiver.
Optical contamination — cleaning is a maintenance schedule, not a heroic act
Every contamination warning we have ever investigated had two things in common: nobody had written down a cleaning interval, and the operator who eventually cleaned it used the nearest rag. Both are fixable.
The scanner already tells you when to clean. Most safety scanners output a contamination percentage or graded warning well before the threshold that trips the OSSD. Log that value weekly for the first two months of operation; you will see the curve characteristic of your environment. Set the cleaning interval at roughly half the time it takes to reach warning, and you will rarely meet the lockout threshold. The naive alternative — cleaning “every Friday” — is either too often (waste) or too rare (lockouts), and almost never matches the actual environment.
On chemicals: use the solvent the manufacturer specifies in the manual. Isopropyl alcohol on a soft, lint-free cloth is the most widely accepted approach for hard-coated optical windows — never apply solvent directly to the window, always to the cloth. Things to keep away from the scanner: ammonia-based glass cleaners (attack coatings), acetone or strong ketones (attack the polymer cover), paper towels and shop rags (abrasive, leave fibres), and compressed air with oil residue. Where spatter is heavy, fit a sacrificial window cover and replace it on a schedule rather than trying to clean the main window every time.
Mounting alignment drift
Drift is the failure mode nobody schedules for. Vibration over time loosens fixture bolts, thermal cycling on an outdoor or high-bay install shifts brackets by fractions of a degree, and a forklift bump nobody reported moves the scanner enough to send the protective field into a static obstacle. The result is a scanner that worked fine for nine months and now trips regularly.
Three habits prevent most of this. First, stiff brackets: two-point mounting top and bottom, mid-span support on long runs, and torqued fasteners with a thread-locking compound rated for vibration. Second, an annual walk-and-verify: with the machine running, the field outline should not flicker on static obstacles at the boundary. Third, store the original field configuration off the scanner (in the safety-controller backup or a controlled file) so you can compare the live field to the as-commissioned one and detect drift in software, not by intuition. For more on the optical and mechanical alignment routine that applies equally to scanners and curtains, see our companion guide on alignment issues and false trips.
OSSD lockout — what triggers it and how to reset safely
The Output Safety Switching Device (OSSD) outputs on a Type 3 scanner carry test pulses — brief low pulses on each OSSD, spaced so the receiving safety controller sees them but the downstream contactor coil ignores them. The scanner checks that each pulse readback matches what it sent. If it does not, the scanner has detected one of:
- A short between OSSD1 and OSSD2 (cross-fault).
- A short of either OSSD to 24 V or 0 V.
- A downstream switching device (contactor, safety relay) that did not respond as expected.
- An internal driver fault inside the scanner.
Any of these latches the scanner into a lockout. The OSSDs stay off, the diagnostic flags the cause, and the unit will not run again until the cause is cleared and a deliberate reset is issued. This is not a nuisance; this is the scanner doing its job. The reset must come from a monitored input, ideally outside the danger zone and visible to whoever issues it, and it must never be wired through software-only logic that can be bypassed. If a lockout repeats within a few cycles after reset, stop. The unit is telling you something, and pressing the button harder is not the answer.
EDM (External Device Monitoring) failures
External Device Monitoring is the loop that gives the safety function its ability to detect a stuck or welded downstream contactor — a requirement for Category 3 architecture under ISO 13849-1. The scanner switches OSSDs; the contactors switch; their mirror (force-guided) contacts feed back into the EDM input. The scanner expects the EDM input to be the logical inverse of the OSSDs within a configured time window (typically a few hundred milliseconds). If it is not, the scanner locks out with an EDM fault.
The recurring problems with EDM are mundane. First, the contactors used do not actually have force-guided mirror contacts and are using auxiliary contacts that can stay consistent with the main contacts during a weld — that is a hardware error caught only by audit. Second, EDM is enabled in the scanner configuration but the feedback wire is missing or shorted — immediate lockout at power-up. Third, the EDM time window is too tight for the actual contactor response time, especially with large contactors or long wiring runs — intermittent lockouts that look random. Fix the configuration to match the hardware, not the other way round.
Scanner-to-scanner interference
Two scanners pointed into a shared volume can see each other’s emitted pulses. The receiving scanner interprets the foreign pulse as a return from a non-existent object — a phantom intrusion appears in the field, the OSSDs drop, and the trip clears as soon as the units fall out of sync. Classic examples are two fixed scanners covering opposite ends of an aisle, a fixed gate scanner facing an AGV-mounted scanner, or two mobile platforms meeting head-on.
Three mitigations, used in combination on busy sites:
- Geometric offset. Mount adjacent scanners at slightly different heights or angles so the active scan planes do not coincide. A 30–50 mm vertical offset is often enough.
- Pulse coding / modulation. Most modern scanners support several pulse signatures or modulation codes. Assign different codes to scanners that share line-of-sight.
- Hardware synchronisation. Where the scanner family supports it, run a sync wire between fixed units so their scans timeshare the field. This is the most robust answer for static installations and is what we recommend on long aisle protection runs.
Wiring faults — the boring half of every lockout
A surprising fraction of stubborn scanner faults trace back to wiring done weeks earlier and never re-verified. The usual suspects:
- Crushed cables in cable carriers — on AGVs and moving gantries, repeated flex eventually breaks a sync or OSSD conductor. A break that intermittently makes contact is the worst kind because it produces random lockouts.
- Shared 0 V return with inductive loads — valves and contactor coils on the same return as the scanner’s 0 V will inject transients into the OSSD circuit. Use a dedicated return for the safety wiring.
- Cable shield grounded at both ends — creates a ground loop and a path for low-frequency hum that can show up as occasional OSSD test mismatches. Ground the shield once, at the controller end.
- Parallel routing with motor power — maintain a minimum 200 mm separation; cross at 90° where they must intersect.
Common error categories (generic, not brand-specific)
Specific error codes vary by manufacturer, but the categories do not. Every safety laser scanner’s diagnostic falls into one of these:
| Category | What it means | Typical root cause |
|---|---|---|
| OSSD fault | Self-test pulses on the safety outputs did not return the expected response | Wiring short to 24V, short to 0V, cross-fault between OSSD1/OSSD2, contactor stuck closed |
| EDM fault | External Device Monitoring feedback did not invert when OSSDs switched | Welded contactor, missing feedback wire, wrong EDM enabled/disabled in config |
| Contamination warning / error | Front window optical attenuation exceeds threshold | Dirty or damaged window, condensation, fine surface scratching from abrasive cleaning |
| Internal hardware fault | Diagnostic on motor, mirror, photodiode or supervisor MCU failed | Bearing wear, impact damage, end-of-life, supply transient |
| Configuration / CRC fault | Parameter set checksum or version mismatch | Aborted parameter download, firmware update interrupted, replacement scanner not re-paired |
| Communications fault | Loss of safe network connection or sync line | Cable damage, EMI on the bus, terminator missing, sync wire open |
Diagnostic decision table
Pin this near the scanner. Maintenance teams who carry a version of this on a laminated card resolve faults in minutes instead of shifts.
| Symptom | Likely cause | Next step |
|---|---|---|
| OSSD drops intermittently, scanner clears within seconds, no lockout | Optical false trip (precipitation, dust, low-reflectivity) | Read intensity/quality log; clean window; check multi-echo config |
| Trips correlate with another scanner cycling in the same aisle | Scanner-to-scanner interference | Apply modulation or sync; offset scan planes vertically |
| Lockout immediately at power-up | Configuration CRC, OSSD wiring fault, or EDM mis-enabled | Read diagnostic code; verify configuration version against drawing |
| Lockout after first machine cycle, EDM error | Contactor feedback not inverting in expected time window | Scope feedback signal; check K1/K2 mirror contacts; verify EDM timeout setting |
| OSSD switches off when drives start | EMI on OSSD or sync wiring, ground loop, supply ripple | Separate from motor cables, ferrites, single-point ground, scope 24V |
| Gradual rise in contamination warning over weeks | Window soiling at rate exceeding cleaning interval | Shorten interval; check window cover integrity; review environment |
| Field intrusion reported with no visible object | Specular reflection, hanging cable, mounting drift | Walk the field with a test object; verify mounting torque; re-teach if needed |
| Persistent contamination error after thorough clean | Window damage (scratching, hazing) or internal optics fault | Inspect window under raking light; if scratched, replace cover or scanner |
Preventive maintenance — a realistic schedule
For a typical indoor industrial install:
- Weekly: visual check of window, log contamination value if accessible, walk the protective field with a 70 mm-equivalent test object at the worst case point.
- Monthly: clean window per manufacturer procedure, verify bracket fasteners are still at torque, review controller event log for recurring warnings that did not become lockouts.
- Quarterly: functional test — verify the safety function (intrusion to safe state to reset) with the machine fully energised, including EDM contactor monitoring.
- Annually: compare live field configuration against the as-commissioned backup; inspect cabling at moving sections; verify response time of the whole chain still meets the ISO 13855 distance calculation.
For outdoor or particularly harsh installs, expect to double the cleaning and inspection frequency. The DAIDISIKE DLD30T-5N perimeter scanner running 40 m fence lines, for example, will see weather contamination an indoor DLD05A3 never encounters, and the maintenance cadence has to reflect that.
When to replace rather than repair
A safety laser scanner is a sealed, type-approved assembly. Field repair beyond replacing the front window cover, mounting hardware and external cabling is not appropriate — any internal change can invalidate the IEC 61496-3 certification. Replace, do not repair, when:
- The diagnostic shows a hardware fault that survives a full power-cycle and a known-good configuration download.
- The contamination warning persists after thorough cleaning of an undamaged window.
- The housing has taken an impact hard enough to move the mirror or motor assembly.
- The unit is approaching the proof-test interval or service lifetime declared in the manufacturer documentation.
Document every swap with serial numbers, the date, and the configuration version restored to the replacement unit. Replacement traceability is a routine audit question under IEC 62046 and the answer should not be “we think it was last winter.”
The bottom line
Safety laser scanners earn their certification by being paranoid on your behalf. The cost of that paranoia is more diagnostic noise than a simpler sensor would generate, and every code is worth reading before it is cleared. If the team treating “press reset, line runs” as the entire maintenance procedure starts treating the diagnostic log as the entire maintenance procedure, both uptime and audit posture improve at the same time. That is, in the end, what the standard intended.
References
- IEC 61496-3 — Safety of machinery: Electro-sensitive protective equipment, Part 3: Particular requirements for active opto-electronic protective devices responsive to diffuse reflection (AOPDDR).
- ISO 13849-1 — Safety of machinery: Safety-related parts of control systems. Fault response, diagnostic coverage and Performance Level architecture for Category 3 and 4 systems.
- IEC 62046 — Safety of machinery: Application of protective equipment to detect the presence of persons. Verification and maintenance requirements.
- ISO 13855 — Positioning of safeguards with respect to the approach speeds of parts of the human body. The safety-distance calculation that ties scanner response time to mounting geometry.