Light Lane
Engineering specification

The published operating envelope: materials, marks, speeds and symbol grades

Six source classes with the physics stated for each, a material compatibility matrix carrying an engineering note on every row, a measured speed model with the five variables that move the ceiling, and the grading standards a symbol is accepted against. The source is a selectable variable: we specify and integrate the source the material dictates, and the parameter set is established on your own material in the capability study. Every figure here is a typical envelope from our own testing, and the binding number is the one written into the acceptance criteria before the build starts.

  • Source selection across fibre at about 1064 nm, pulse-controlled MOPA, green at 515 or 532 nm, UV at 355 nm, CO2 at 9.3, 10.2 or 10.6 micron, and ultrafast picosecond and femtosecond, chosen against what the material does with the energy
  • A typical timber brand held across 40 to 120 m/min on a single galvo head, a full height dense brand topping out around 75 to 85 m/min on one head, and the five variables that move that ceiling published beside the figures
  • Data Matrix generated to ISO/IEC 16022 and graded to ISO/IEC 15415, linear symbologies graded to ISO/IEC 15416, accepted at the threshold the scanner class your operation carries supports
Send us a partBook a site scope

The source is a selectable variable

For twenty years the laser was the product and the source decided what you could mark. The source is now a selectable variable, and almost everything the category still publishes as a property of laser marking is a property of one wavelength somebody chose without thinking about it.

We specify and integrate the source the material dictates, and the parameter set is established on your own material in the capability study. The matrix below is a selection matrix. It states what the application requires, and that requirement is settled on measured results before anything is bought.

What decides the answer is what the material does with the energy at that wavelength. Absorption in copper at 1064 nm is a few percent at room temperature and climbs sharply below about 600 nm. PET barely absorbs 10.6 micron and absorbs 9.3 micron efficiently, which is the difference between a code you can read and no code at all. UV at 355 nm breaks the molecular bond photochemically rather than heating it, so there is effectively no heat affected zone. Ultrafast pulses remove material by ionisation rather than melting. These are properties of wavelengths, published in the optics literature and checkable independently of anything on this page.

Get the source selection wrong and no amount of power, software or setup time recovers it. That is why the selection is made in the capability study, on your own material, before capital is committed.

Source selection: what the application requires

One row is selected per application, against the material, the mark and the conditions the mark has to survive. The source is specified and integrated to what the material dictates, and the parameters are established on your own material in the capability study.

Source class What the physics does Where the specification selects it
Fibre at about 1064 nm Absorbed by metal, producing annealing, an oxide colour change or material removal depending on power and pulse settings. Steel, stainless, aluminium, titanium and engineering alloys, anodised and coated metal, and some filled or pigmented plastics.
MOPA pulse-controlled fibre Pulse generation is separated from amplification, so pulse duration, repetition rate and pulse energy move as three independent variables, typically 4 to 200 ns at 1.6 kHz to 1 MHz. Deep black formed inside an anodised layer, annealed marks on stainless with no material removed, low heat input on thin and coated stock, and clean marks on plastics that foam or char under long pulses.
Green at 515 or 532 nm Absorption in reflective metals is only a few percent at 1064 nm at room temperature and climbs sharply below about 600 nm. Copper, brass, gold and silver, where the process needs the beam to couple into the metal in a stable, predictable way rather than reflect off it.
UV at 355 nm Breaks the molecular bond photochemically rather than heating it, so there is effectively no heat affected zone. Glass, quartz, ceramics, silicon, PCBs and heat sensitive plastics including PP, POM, PVC, PBT and polyamide.
CO2 at 9.3, 10.2 or 10.6 micron Absorbed by organic material, producing localised burning, charring, vaporisation or removal of a top coating. PET barely absorbs 10.6 micron and absorbs 9.3 micron efficiently. Timber and wood products, paper, cartonboard and corrugated, acrylic, many plastics, leather, glass and ceramics, and painted or coated surfaces.
Ultrafast picosecond and femtosecond Material is removed by ionisation rather than by melting, so no heat affected zone is introduced into the part. Brittle, transparent and thin-film work where a heat affected zone is itself the defect, and high contrast black on stainless that survives passivation and autoclaving. Industrialised equipment now rather than laboratory equipment.

Pulse width control turns one head into several processes

A standard Q-switched fibre laser has a fixed pulse width that stretches and weakens as the frequency changes. A MOPA source separates pulse generation from amplification, so pulse duration, repetition rate and pulse energy move independently. That is not a bigger laser. It is a different set of processes out of the same head.

Black marking on anodised aluminium. The short pulse changes the anodic layer rather than cutting it, so the mark forms inside the coating, the surface stays sealed and the finish stays intact. Published manufacturer application note parameters on a 20 W MOPA with a 254 mm lens run around 8 ns at 200 kHz, 1500 mm/s, single pass, with line spacing setting the tone from grey through to a dense black. Those are the manufacturer's figures, attributed to their application note rather than measured by us.

Annealed marking on stainless with no material removed. Pulse control puts in enough energy to grow a sub-surface oxide and stops before ablation, so the mark is a colour change with no depth and the chromium oxide passivation layer that gives stainless its corrosion resistance stays intact. That is why it is the accepted method for surgical instrument identification, and a published manufacturer durability test found laser-marked codes surviving at least 500 sterilisation and cleaning cycles. Their test, cited as theirs.

Reduced heat affected zone on thin, coated and sensitive stock: less warping, less discolouration, less micro-cracking, no melted edges. And plastics marked without foaming or charring, by dropping to a few nanoseconds.

Colour on stainless is thin-film interference. The laser grows a nanometre-scale oxide film and the colour comes from the reflection off the top of the oxide meeting the reflection off the oxide to metal interface. Published film thicknesses sit in the tens to hundreds of nanometres, chromium-rich oxides give greens and blues, iron-rich oxides give browns and reds, and the result is driven mainly by pulse energy and spot overlap, which is exactly why pulse control is the enabling variable. It is a temper colour palette rather than a colour space, golds, bronzes, browns, blues, purples and some greens, and it shifts with viewing angle because interference does. Published research states directly that the reproducibility of the markings and the processing times often do not meet the requirements of industrial manufacturing processes, and that the process is still not fully understood because of the number of mutually influencing parameters. So colour is specified as an identity and decorative capability with a defined envelope, oxide, anneal or ablation contrast is specified wherever the mark has to be graded and read, and durability under abrasion, heat and chemistry is characterised on your own part in the capability study rather than assumed from public data.

Fibre and CO2 at close range

These two carry most industrial marking between them, and they are not interchangeable. The wavelength decides what the material does with the energy, which decides whether there is a usable mark at all.

Property Fibre, about 1064 nm CO2, about 10.6 micron
Marks well Steel, stainless, aluminium, titanium, most engineering alloys, anodised and coated metal, and some filled or pigmented plastics Timber and wood products, paper, cartonboard and corrugated, acrylic, many plastics, leather, glass and ceramics, and painted or coated surfaces
Where the other source is specified instead Bare timber, paper and card, clear glass and many unpigmented plastics, which are CO2 or UV work Bare metal, which is fibre, MOPA or green work
How the mark forms Absorbed by metal, producing annealing, an oxide colour change, or material removal, depending on power and pulse settings Absorbed by organic material, producing localised burning, charring, vaporisation or removal of a top coating
Typical appearance Dark grey to black on steel and stainless, bright substrate showing through on anodised aluminium, engraved relief where depth is specified Brown to black char on timber, high contrast where a printed or coated top layer is removed on card, frosted white on glass
Common industrial use Serial numbers, Data Matrix, part numbers and asset marks on components, tooling and equipment Grade stamps, brands, batch codes and Data Matrix on timber, packaging and coated product
What the specification does with the edge cases Copper, brass and other reflective alloys absorb little at this wavelength, so a green source at 515 or 532 nm is specified where the mark has to land in a reflective metal, and the cycle time comes off the mark made on your actual alloy CO2 cannot mark bare metal, because metal reflects far more of a 9 to 11 micron beam than it absorbs, so bare metal is fibre, MOPA or green work. On thin card the window between a clean mark and burn through is narrow, so power is set against the actual board grade you buy and a board change is a re-test

Material compatibility, with an engineering note on every row

Read the notes column as carefully as the rest of the row. The source column states what the application calls for, and where a material family is genuinely variable the note says test, because a measurement on your own stock is worth more than a rating that would not hold across it.

Material Source specified What the mark looks like What it survives Engineering notes
Stainless steel Fibre at about 1064 nm, or MOPA where an annealed mark with no depth is specified Annealed dark oxide with no material removed, or an engraved mark where depth is specified Wash-down, handling, weather, most cleaning regimes An annealed mark leaves the passive surface intact, which matters on food contact and marine parts. An engraved mark breaks that surface and can become a corrosion site, so the mark type is a real decision, not a setting.
Stainless steel, colour marking MOPA pulse-controlled fibre Temper colours grown as a nanometre-scale oxide film: golds, bronzes, browns, blues, purples and some greens Handling and service, characterised on your own part rather than assumed from public data The colour is thin-film interference between the reflection off the top of the oxide and the reflection off the oxide to metal interface, driven mainly by pulse energy and spot overlap, so it shifts with viewing angle by physics. It is specified as an identity and decorative capability with a defined envelope, and oxide, anneal or ablation contrast is specified wherever the mark has to be graded and read.
Mild and carbon steel Fibre at about 1064 nm Dark oxide or engraved relief, good contrast on a clean surface Handling and normal service, provided the part is coated or oiled as it would be anyway Durability is governed by the steel rather than by the mark. Bare mild steel corrodes over the mark like it corrodes everywhere else. Where the process allows it, mark before coating.
Aluminium, bare, extruded or machined Fibre at about 1064 nm Grey to dark grey, noticeably lower contrast than stainless Handling, weather, the service life of the part Contrast depends on alloy and surface finish. A brushed face, an as-extruded face and a machined face all read differently. Test on the finish you actually run, not on a sample coupon.
Anodised aluminium Fibre, or MOPA where a dense black formed inside the layer is specified Very high contrast: the dyed anodic layer is removed and the bright substrate shows through, or a dense black formed inside the coating with the surface left sealed Handling and weather, for as long as the anodic layer itself lasts The mark exists only within the coating thickness, so sustained abrasion will take it off. Within that limit it is fast, clean and one of the most repeatable marks we run.
Brass, copper and other reflective alloys Green at 515 or 532 nm A clean, repeatable mark where the beam couples into the metal rather than reflecting off it Handling, weather and the service life of the part Copper, brass, gold and silver reflect infrared. Absorption climbs sharply below about 600 nm, so a green source at 515 or 532 nm couples into these metals in a stable, predictable way. Reflective alloys are a source selection question, and the cycle time is quoted off the mark made on your actual alloy in the capability study.
Painted, powder coated and coated metal Fibre or CO2 The coating is removed to expose the substrate, which gives strong contrast As long as the exposed substrate is protected or is not corrosion prone Think about this one twice. Removing a protective coating on outdoor or marine steel exposes bare metal at the mark. Either mark before coating, or plan a re-seal, or accept the exposure knowingly. Decide it at scoping, not at commissioning.
Dried and kiln dried timber, including pine CO2 Brown to black char, sharp edged on a planed face, softer edged on rough sawn Stacking, handling, weather, yard time and transport The core timber application and a well proven one. Contrast varies with species, resin content, surface finish and moisture. Rough sawn faces set how small a code can go before the module edges blur.
Green and high moisture timber CO2 Lower contrast and more variable than on dried stock Handling and yard conditions once the surface dries Surface moisture absorbs energy and escaping steam disturbs the mark as it forms. It does mark. The number that matters is the one measured on your material at your moisture content, established in the capability study and written into the acceptance criteria.
LVL, glulam and laminated timber CO2 The same char mark as solid timber, with a visible tone change where the mark crosses a glue line Handling, weather and structural service Glue lines and the wood either side do not burn the same way. For a human readable brand that is cosmetic. For a Data Matrix crossing a glue line it can cost you symbol grade, so codes get placed deliberately rather than dropped in the middle.
Treated timber, including H3 and H4 CO2 Char mark, with the colour shifted by the treatment chemistry Outdoor exposure and yard handling, as the treated timber itself does The treatment changes both the appearance of the mark and the fume that comes off it. Extraction and the safety data for your specific treatment need reviewing for your product, never assumed from another mill.
Engineering plastics: ABS, polycarbonate, acetal, nylon CO2, MOPA on filled or pigmented grades, UV at 355 nm on heat sensitive grades A foamed light mark, a carbonised dark mark, or a colour change, depending on the polymer and its additive package Handling, cleaning and service life, subject to which mark type is achieved Plastics are the least predictable family on this page. Pigment and additives drive the result more than the base polymer does, and a resin supplier change can change your mark. Always tested, always on your actual grade. Dropping to a few nanoseconds on a MOPA source is what stops a grade foaming or charring where a long pulse would.
Polypropylene and polyethylene, unfilled UV at 355 nm, or fibre where the resin carries a marking additive A mark formed by breaking bonds at 355 nm, or a foamed or contrasting mark at 1064 nm where an additive is in the resin Handling, cleaning and service life, subject to the contrast achieved The hardest common polymers to mark. At 1064 nm they usually need a marking additive in the resin. At 355 nm, with parameters set against the actual grade, they can be marked without one. Contrast is the variable rather than whether a mark forms, so this is a sample and test conversation, and the grade you actually buy is the grade we test.
Composites and laminates Usually CO2, with UV where the surface layer is heat sensitive Governed by the surface layer, not by the core Whatever the surface layer survives The mark is a reaction of whatever is on the outside: gelcoat, veneer, melamine, film. Two panels with the same core and a different surface behave completely differently. If you run both, send both.
Cartonboard, corrugated and kraft CO2 The printed or coated top layer is removed to reveal the fibre underneath, which reads as high contrast Warehouse handling, transport and moderate humidity Fast, and with no consumable behind it. The window between a clean mark and burn through is narrow on thin stock, so power is set against the actual board grade you buy, and a board change is a re-test.
Coated films and flexible packaging CO2 at 9.3 micron on PET, or MOPA where heat input has to stay low Lacquer or coating ablated to reveal the layer beneath, or a code formed in a laser reactive coating Depends on the laminate structure PET barely absorbs 10.6 micron and absorbs 9.3 micron efficiently, so the wavelength decides whether there is a readable code at all. Barrier integrity is the governing question on an unsupported film: the laminate structure and the ablation depth are specified together, and both are tested on your own film before anything is committed.
Film, foil and board carrying a laser reactive coating CO2 or fibre, selected against the coating chemistry High contrast durable codes including clear to white, formed in a coating laid down at the converter by flexo or gravure, usually as a patch Handling, transport and the pack's own service conditions The coating carries the reaction, so the substrate question moves off the line and onto the converter. There is no ink and no consumable at the marking station, and the patch can be printed under a barrier laminate so the code is formed inside the pack and reads as tamper evident. Coating chemistry and patch placement are specified with the converter before the station is designed.
Glass and ceramics UV at 355 nm or ultrafast where the part has to stay intact, CO2 where a frosted mark suits A clean mark with essentially no heat affected zone at 355 nm or under ultrafast pulses, or a frosted white mark produced by controlled surface fracture under CO2 Washing, handling and normal use A CO2 frost mark is micro-fracture, which introduces a surface flaw. On containers under internal pressure, thermal cycling, or with a tight strength requirement, that is assessed rather than assumed away, and UV or ultrafast is specified where the strength case rules the flaw out.

Beam time is path length, not area

A single galvo head holds a typical timber brand across roughly 40 to 120 m/min. Across that range the mark stays legible, stays repeatable and lands where it is meant to, and it covers most of the line speeds we see in New Zealand timber processing.

A full height dense brand, around 40 mm high and up to roughly 600 mm long, tops out around 75 to 85 m/min on one head, because the ceiling is how much stroke length one scanning mirror pair can draw in the time the part spends in its field of view, so above that the specification is a second head sharing the mark on the same encoder.

That ceiling moves. It is the outcome of five variables, mark height, character density, total mark length, line speed and available power, and the table below sets out what each one does to the number and what sizing it takes from you.

Two things follow. The right question at scoping is not what speed a laser runs, it is what mark you need at what speed, because those two together decide the head count and most of the cost. And a published speed with no mark stated beside it is decoration, so ask anyone quoting one exactly what we would ask: at what mark height, at what character density, over what mark length, at what power.

Published figures are typical envelopes from our own testing. The binding number is the one written into the acceptance criteria after the capability study on your own material.

The five variables that move the ceiling

Change any one of these and the maximum line speed changes with it, which is exactly why a single published speed figure carries no information without the mark stated alongside it.

Variable Effect on maximum line speed What we need from you
Mark height Taller characters take longer per stroke and eat more of the available marking window. A full height brand is the heaviest case on the line. The actual artwork or grade stamp at the size it has to appear on the product.
Character density Total stroke length is what costs time, not character count. A dense grade stamp with fine detail costs far more than the same footprint of open text. The real mark, not a tidied up version of it. Simplified artwork produces a speed figure you cannot use.
Total mark length A longer mark needs a longer window at the same speed, and the window is bounded by the head field and the spacing between products. Mark length, product length and the gap between products as the line actually runs.
Line speed The product is moving while the mark forms, so everything above is judged against the time the part is genuinely in front of the head. Real line speed, including the fast setting you use when you are catching up, not the nominal one on the drawing.
Available laser power More power lets the same mark form in less time, up to the point where the material governs the result rather than the laser. The material, so the point where extra power stops helping and the substrate takes over is found on your own stock.
Head count, the answer above the ceiling Two heads sharing one mark on the same encoder is the standard route past the single head ceiling, and they compose one mark between them rather than two. Confirmation of the fastest mark you actually need to hold, so heads are sized once at scoping.

Field size, focal length and part presentation

The optics decide how much of the part one head reaches and how much variation the part is allowed to arrive with. Both are numbers with an engineering answer attached, and both are settled at survey rather than discovered at commissioning.

What the specification sets What decides it The engineering answer
Marking field from one head The scan lens. Published manufacturer marking windows run from about 31 by 38 mm up to about 601 by 440 mm, across focal lengths from 64 mm to 600 mm. On a wide web or a multi-lane line that is the difference between one head and four, so the lens is selected against lane width and mark size, with spot size and available power checked against it in the same pass.
Depth of field Focal length, spot size and the height variation the part actually arrives with. Three axis dynamic focus holds focus across height, taper, curvature and inclined faces, and a distance sensor corrects continuously. Where the variation runs wider than the optics hold economically, fixturing takes the remainder, and that split is priced at survey.
Part position and rotation How the part presents at the head: lateral drift, rotation, and the gap between products. Vision corrects X, Y and rotational offset before the mark is placed. Presentation variability is an engineering budget with a number attached: the height and orientation variation held is stated, along with the point where fixturing is cheaper than optics.
Marking on the fly Line velocity, taken from the encoder the line already has. The mark geometry is compensated for line velocity so a shape drawn onto a moving surface comes out true rather than sheared. A real encoder signal is what makes that exact. Nominal line speed off a specification sheet is a different number, and marking on the fly shows you the difference.

Mark types, and what each one is for

The right mark is decided by who or what reads it, and under what conditions. Minimum practical sizes are governed by the substrate, the reader and the read distance, so this table states what determines the number and the capability study returns the number itself.

Mark type What it is Best used for What sets the size
Data Matrix A 2D matrix symbology defined by ISO/IEC 16022. The ECC 200 version carries Reed Solomon error correction, so a partially damaged symbol still decodes. Small parts, dirty or damaged surfaces, and anything that has to keep reading after handling Denser than QR for short payloads, which is why it dominates direct part marking. The minimum is set by the smallest module your substrate holds cleanly and by the reader optics and working distance, and it is established by marking your material and reading it with the intended scanner.
QR Code A 2D symbology almost any phone camera reads, with selectable error correction levels Marks a person needs to scan without special equipment, or a code that carries a link as well as an identifier Larger than a Data Matrix for the same short payload at the same module size. Choose it where phone readability matters more than footprint, and Data Matrix where footprint matters more.
Human readable text Characters an operator, inspector or customer reads by eye Grade, treatment, mill or plant identity, dates, and anything a person acts on without a scanner Character height is driven by read distance and lighting rather than by the laser. Tell us where it gets read from and under what light, and it is sized against your substrate.
Serial numbers A unique identifier per item that ties the physical part to its record Warranty, service history, asset registers and certification traceability The sequence is owned by one system, and that system is your ERP or MES rather than the marking software. The Control Layer receives or requests the number and writes the mark confirmation back against it.
Batch and lot codes An identifier shared across a production run Recall containment, and the one-up one-down traceability the Food Act 2014 requires, which is outcome based and mandates no particular format The granularity of the code is the granularity of your recall. A narrower batch code costs nothing more to mark and can cost a great deal less to act on.
Grade stamps and brands The full mill or plant mark: grade, treatment, standard reference and producer identity Timber and wood products, where the mark is a claim about the product rather than just an identifier The heaviest mark type on a moving line, and the one that sets the throughput ceiling published above. Size and content are usually fixed by the scheme or the customer, so line speed and head count become the variables that move.
Logos and brand marks Vector artwork marked as artwork Product branding and customer facing surfaces, where the mark is part of how the product looks Judged by eye rather than by a scanner, so the governing factor is edge definition on the substrate rather than a symbol grade. Fine detail is the first thing lost on rough sawn or coarse surfaces.
Variable date and time fields Fields composed at the moment of marking, from the line clock or from a field in your system Pack date, best before, shift and run codes Composed at mark time rather than pre-rendered, so a date rollover or a shift change needs no job change. Where a system owns the value, that system stays the source of truth.

Symbol grading and the standards written into acceptance

A code either decodes or it does not, and a customer or an auditor may want more than that: a graded measurement of how well it decodes, against a published standard. Every standard below is a public document you can buy and check independently of anything on this page, and each one lands inside acceptance criteria as a threshold with a test reference beside it rather than in a badge row.

Standard What it covers How it is used
ISO/IEC 16022 The Data Matrix symbology specification, including the ECC 200 error correction used in industry Data Matrix is generated to it, and it is the reference point when a customer specifies Data Matrix rather than just asking for a 2D code
ISO/IEC 15415 Print quality grading for 2D symbols, producing a grade from a set of measured parameters rather than a pass or fail The grading standard for 2D marks, applied where a supply agreement or an audit specifies a minimum symbol grade rather than just that the code reads
ISO/IEC 15416 Print quality grading for linear, one dimensional barcodes The grading standard where legacy 1D codes still run in the process alongside 2D
MIL-STD-130 The United States Department of Defense standard for identification marking of military property, including item unique identification carried in a machine readable mark Specified to and verified against where your operation works to it, in defence supply chains and in commercial parts that end up inside one
NASA-STD-6002 Applying Data Matrix identification symbols on aerospace parts Named where aerospace direct part marking requirements are passed down through a prime contractor
AS9132 Quality requirements for Data Matrix direct part marking in aerospace Specified to and verified against where your operation works to it, usually named in the same breath as NASA-STD-6002
Verification in practice Measuring a marked symbol with a verifier against 15415 or 15416, which is what produces a grade you can put in front of an auditor. A hand scanner reading the code is a different class of evidence. Tell us the grade and the standard your customer requires at scoping, and the verification method, who performs it and the accept threshold are written into the acceptance criteria before the build starts
What an acceptance criterion looks like For example: Data Matrix legible after 90 degree caustic wash, verified per ISO/IEC 15415, accepted at grade C or better after 30 cycles on the scanner class your operation carries. Every requirement of this shape is mapped to the named test that closes it, so the grade is a measured result on your own product rather than a promise made in advance

How a standard is carried into your acceptance criteria

Confirmed

  • Light Lane specifies to, designs to, verifies against and supplies handover documentation for ISO/IEC 16022, ISO/IEC 15415, ISO/IEC 15416, MIL-STD-130 and AS9132 where your operation works to them.
  • Data Matrix is generated to ISO/IEC 16022 and graded to ISO/IEC 15415, linear symbologies are graded to ISO/IEC 15416, and the accept threshold is set against the scanner class your operation actually carries.
  • Every standard named on this page is a public document. Buy it, read it, and check this page against it.

Not confirmed

  • A symbol grade is a measurement rather than a specification, so it is established on your substrate with your reader class during the capability study, then written into the acceptance criteria as a threshold with the test that closes it.
  • A grade holds against the surface condition it was measured on, so the acceptance criteria state the condition alongside the threshold, and re-verification intervals are set where the surface changes across a year.
  • The evidence is the mark on your own material: the sample, the parameter set that produced it and the verifier reading, which is exactly what the Marking Process Capability Study returns.
Send us a part See the delivery sequence

The controller-aware core, and what it emits

The desktop application and the industrial marking layer run on the same controller-aware core, written in Nelson. The figures below are from that core. One internal, dialect-neutral motion program is emitted into each path, and power scaling alone is four quantisations of the same physical intent: an internal 0 to 10000 range mapped onto the GRBL $30 value read live off the controller at connect time, the Marlin 0 to 255 integer, Smoothieware 0 to 1.0 at four decimal places, and Generic 0 to 1000.

Controller path Status Notes
GRBL Fully supported Controller-aware generation: the machine's own $30 and $31 registers are read at connect time and every emitted power value is scaled to that machine's real range, with $110 and $111 queried to derive and validate the effective maximum feed rate. Arc fitting is labelled Experimental in the product, so it is selected deliberately on suitable jobs rather than run as the default.
Marlin Fully supported For Marlin based machine workflows, with power quantised to the 0 to 255 integer range the dialect defines.
Smoothieware Fully supported For Smoothieware based setups, with power emitted from 0 to 1.0 at four decimal places.
Generic or custom G-code Supported For machines that fit a broader G-code workflow, with power emitted across a 0 to 1000 range.
Ruida Available in alpha An external bridge, labelled alpha: an active support path being evaluated and improved rather than one at parity with the paths above. Validate coverage of your own Ruida workflow edge cases before you rely on them.
Regression discipline across the dialects Held in continuous integration A golden fixture suite holds emitted machine output stable across three dialects, 42 fixtures across two images and seven processing variants, with a conformance validator checking output against the dialect rules. A change to the geometry engine shows up as a failed fixture rather than as a different part on the floor, which is the answer to the question an industrial buyer rarely gets to ask: what happens to my line when the software updates.

Integration interfaces published as part of the specification

The Light Lane Control Layer sits between the plant's control and business systems and the marking source. It reads what the plant already knows, composes the mark for the part at the head, fires it, and writes the completion back against the job. Each pattern carries its own authentication model and its own documented failure edge, and the pattern is chosen with your IT team against what your environment already runs. The full architecture is on the integration page.

Interface What it carries Authentication model and what the plant provides
PLC digital I/O Trigger, ready, fault and part present signals. The simplest and most robust way to start a mark. Spare I/O, a trigger source such as a photo eye or an encoder, and agreement on signal levels with whoever owns the PLC program. The posture is read only: the agreed signals, with ladder logic and function blocks left untouched, and any write scoped as a specific handshake and signed off separately.
PLC over ethernet or fieldbus Product and job data rather than just a trigger: which product is running, which grade, which batch, which length. A tag list and a named contact who owns the program. OPC UA runs certificate-based security where the controller carries it, and Modbus TCP, which carries no authentication of its own, is scoped to a network segment agreed with your control engineer.
ERP and MES over REST or SOAP The job, the batch, the serial and the identifier itself. Reads run against NetSuite, Infor M3, Microsoft Dynamics 365, SAP, Cin7, Unleashed and Fishbowl, and against the WMS, TMS, CMMS, calibration management and yard management systems beside them. API key or OAuth client credentials, read access to the fields the mark is composed from, write access scoped to the single field carrying mark confirmation, and a test environment so the first integration test runs off the live line. No personal staff login is requested or wanted.
Read-only database view or watched file drop The same job and identifier data, where a system carries no usable API. Older systems are common in real plants and this path runs properly. A least-privilege service account restricted to one read-only view, or a scoped share account on the file drop, with a location, a format and a named owner. Both carry a defined failure edge and are specified in writing like every other pattern.
Message queue or sanctioned middleware Job and completion messages through the broker or middleware your organisation already runs. Delivery is asynchronous, so the job data is resident at the head before the part reaches the field of view, and the composition time in the envelope above stays a marking figure rather than a network figure. The broker or middleware endpoint, the credential your platform team issues, and the topic or queue naming your own standards define.
Encoder and line speed feed True line velocity, so the geometry is compensated and the mark lands true on a product that is moving while it forms. A real encoder signal where one exists, with its resolution and mounting point confirmed at survey.
Confirmation written back What was marked, on which part, when, by which machine and station, written as an audit record against the job. Agreement on which field carries confirmation and what counts as confirmation in your process, so the loop closes without a second database to maintain.
Vision and read back The mark read after it is made, so a suspect mark is caught at the head rather than at dispatch. A decision on hold or divert, chosen by you and written into the acceptance specification before the build, along with what the operator sees and what the record shows.
Network path The connection runs outbound from the marking station to your endpoint over TLS. Integrating a marking line requires no inbound path into the plant, and on an isolated OT network the station runs air gapped. Your IT and OT policy, taken at scoping. Egress to one endpoint on one port is the whole ask, and your business system stays unreachable from the public internet without changing anything about how the line marks.

The data round trip, and what the station does when a system underneath it moves

Your system owns the identifier. The Control Layer composes the mark from it, fires it, and writes the completion back. Every behaviour below is written into the interface specification before the build starts, and a record-keeping problem is never allowed to become a production problem.

  1. Step 1

    Your system issues the identifier

    The ERP, MES or PLC already running the business decides what this part is: the job, the batch, the grade, the serial. That system stays the source of truth, and adding a marking head changes nothing about who owns it.

  2. Step 2

    The Control Layer composes the mark for the part at the head

    The PLC tag identifying the physical part is read, the job and batch are pulled from the record that already exists, and a fresh mark is composed against that part's grade and dimension: fixed logo, standards text, the variable date, time, shift, run and pack fields, and a machine readable code. Nothing is pre-rendered and nothing is stored twice.

  3. Step 3

    The head fires inside the window the part gives you

    An encoder tracks the part, the geometry is compensated for line velocity, and the scanning mirror pair draws all of the mark before the part leaves the field of view. Three systems agree inside milliseconds: the control system that knows what the part is, the business system that knows what it should say, and the optics that have to finish drawing in time.

  4. Step 4

    The completion is written back against the job

    Timestamp, operator, machine and station, written as an audit record into the single field scoped for it. That is the record a traceability process actually needs, and it closes without anyone rekeying anything.

  5. Step 5

    Through an outage the station runs from cache and the records queue

    The station fills a pull-ahead cache rather than calling the business system per part, so a scheduled maintenance window never stops a physical line. Completion records queue locally and replay in order, carrying the original mark timestamp rather than the write time, which is the difference between an audit trail and a fiction. Cache depth and hold duration are sized against your run length and line rate at the specification stage, then written into the acceptance criteria as a number.

  6. Step 6

    When the data underneath moves, the station holds rather than inventing

    If the cache runs dry the station stops issuing marks rather than composing from stale data. It never invents an identifier, never substitutes a default, never writes a placeholder and never increments a serial itself. If the PLC signal is missing, or two signals disagree about the part at the head, it refuses to match and logs the exact signal state at that instant, so your control engineer debugs from a record rather than from a description. If a schema or a credential moves underneath the integration, it fails the pull rather than composing a mark from fields it no longer understands. Hold or divert is chosen by you and written into the acceptance specification before the build starts.

Environment, installation and site services

Much of this is equipment dependent, so the table states what sets each number and where it is fixed. A figure that belongs to a specific source, enclosure and duty belongs to that installation, so it is stated with the design and confirmed in writing rather than carried over from another site.

Item What matters Where the number comes from
Fume extraction Marking produces fume and particulate. Timber, treated timber, coatings and plastics each produce different fume, and treated timber needs the safety data for the specific treatment reviewed rather than assumed. Set by the material, the mark rate and the enclosure. The airflow figure is specified against your material, including the safety data for any treatment on it, and stated with the design.
Enclosure, interlocks and the safety circuit Class 4 laser safety is enclosure, interlocks, beam containment and a rated safety circuit. An inline head sits inside an enclosure with interlocked access, so the machine is safe to work beside during normal running and safe to open for service. Designed into the install against ISO 11553-1, IEC 60825-1, IEC 60204-1 and ISO 13849-1. The safety class belongs to the installed arrangement, so it is stated for your installation once the source and the enclosure are fixed.
Ingress and wash-down Food, meat and seafood plants wash down with water, heat and chemistry. That is an enclosure, seal and cable gland question, and it is answered before the build rather than after the first wash. Set by your wash-down regime and chemistry. The rating is agreed with you before the build, against how that area is actually cleaned rather than against a plant wide assumption.
Ambient temperature and humidity Sources, optics and control electronics all have operating ranges. A chilled room, a kiln exit and an open yard are three different problems. Set by the specific source and enclosure selected, and checked against your actual location, including the worst week of the year rather than the annual average.
Dust and airborne fibre Sawdust and airborne fibre settle on optics, and a dirty lens changes the mark well before it stops the mark. Set by your environment. Positive pressure, air knives and a cleaning interval are all normal answers, and the interval is established on your line rather than copied from a manual.
Vibration and mounting A galvo head marking a moving product needs a stable mount. Vibration from a conveyor, a saw or a chipper shows up in the mark before it shows up anywhere else. Set by the structure available at the marking point. This is one of the things a site visit exists for.
Electrical supply Supply, phase and protection for the source, the chiller where one is fitted, the extraction and the control cabinet. Set by the equipment selected, confirmed in writing once the source, the chiller and the extraction are settled, and checked against what is actually available at the marking point.
Compressed air Used for optic protection and for some assist and cleaning arrangements where they are fitted. Set by the configuration, and specified against your available supply rather than against a plant standard that may not exist.
Network A path from the marking station to the PLC, ERP or MES, and a decision on whether that path crosses the plant network or stays on its own segment. Set by your IT and OT policy. The station is specified to it, outbound only over TLS, and the policy is taken at scoping so the design carries it from the start.
Floor space and access Room for the enclosure, extraction, cabinet and service access, plus the route to get all of it through your doors. Set on the site visit. Door widths and overhead clearance have changed more designs than any written specification ever has.

Method selection: when the specification calls for something other than laser

The selection is made in the capability study, with the physics stated and the alternative named. Where the physics routes the job to a coating, a printed panel, a plate or a label, the numbered report says so and states the reason, so a buying group holds an engineering result before capital is committed.

What the application requires What the physics dictates What we specify
An arbitrary brand colour or a Pantone match A laser deposits nothing. Every laser mark is the substrate reacting to energy: oxide, anneal, char, foam, frost or an exposed underlayer. A laser reactive coating laid down at the converter, or a printed panel the laser then marks into. That is a printing requirement, and we specify it as one.
A pale code on a dark polymer pack A foaming mark is the opposite of a char mark. The laser degrades a foaming agent in the polymer, the pores scatter light, and the result reads as a pale mark on a dark background. It depends on the resin and its additive package. A material conversation before it is a laser conversation. We test the actual resin, and where the additive package will not carry the contrast we specify a laser reactive patch or a printed panel and say so at the scope.
Small character coding at very high speed with a mandatory ink colour Throughput is bounded by drawing time rather than by the head, so a two-line date code and a full height dense timber brand are the same physics at different path lengths. The same manufacturer publishes 15 m/s, which is 900 m/min, on a 60 W CO2 laser coder, and roughly 293 m/min on its fastest continuous inkjet printer. Where a specific ink colour is mandatory, or the substrate stays unreactive after test, we specify continuous inkjet and say so at the scope. Where the driver is consumable spend and uptime on a substrate that reacts, we specify the laser and publish the mark that goes with the speed.
A surface that must carry no modification at all Laser marking modifies the surface permanently. Where any modification is disallowed by a warranty, a coating specification or a sealed food contact surface, that alteration is itself the defect. A tag, a plate or a label carrying the identifier, tied to the same record. Marking a part and identifying a part are two jobs, and an attached identifier is sometimes the correct engineering answer.
A line running a few hours a week against a small consumable spend This one is commercial rather than physical. A laser removes a consumable cost, so the return is bounded by the consumable spend it removes. If the consumable spend is small and the line runs a few hours a week, there is no payback, and no source selection changes that. We do the arithmetic with you and show you the number that says stay with what you have. That takes an hour and saves everyone a quarter.
Part presentation wider than the optics hold economically Marking on the fly needs the surface inside the focal depth, at a predictable orientation, at a predictable moment. Dynamic focus and vision widen that window substantially, with a number attached to how far. Fixturing takes the variation beyond what the optics hold, or the mark moves to a point in the process where the part is controlled. The split between optics and fixturing is priced at survey, and where neither is economic the identifier goes on a plate or a tag instead.

How every number on this page becomes a number on your material

One route closes every open value here: the Marking Process Capability Study, run on your own material, with the process you would actually run, before anything is committed.

  1. Step 1

    Send the part or the material

    The real thing, in the condition it will be in when it gets marked. Green rather than dried if green is what runs. The rough sawn face rather than the planed one. Both finishes if you run both. Nearly every failed marking project was decided in a meeting room with no part in the room.

  2. Step 2

    It is marked with the process you would actually run

    Same source class, same mark, same artwork at the size it has to appear on the product. What comes back is the result production gives you, rather than a demonstration piece marked slowly on a good day.

  3. Step 3

    You get the part back with a numbered report

    The report states the substrate and surface condition as received, the parameter set, the verification grade against the applicable standard, the durability result against the conditions you name, and the determined process window, including how much margin that parameter set carries. It is the first checkable object in the programme.

  4. Step 4

    Where it has to run at speed, the speed comes with the mark attached

    Mark height, character density, mark length, line speed and available power together decide the answer, and you get that answer with the ceiling attached, in the same form as the timber figures above, including the point where a second head sharing the mark on the same encoder is the specification.

  5. Step 5

    The numbers go into the acceptance specification, before the build

    What the system has to achieve is written down and agreed by both sides before anything is built, with every requirement mapped to the named test that closes it, and staged payment released against that document. What goes in is the set of numbers on this page, made specific to your product: the symbol grade with the verifier, the standard and the accept threshold, the speed figure with the mark height, character density and mark length it was measured at, and the process window with its margin. The sequence that document sits inside is set out in how a programme is delivered.

Engineering questions, answered with numbers

Can you give a line speed over the phone?

For a typical timber brand, yes: roughly 40 to 120 m/min on a single galvo head, with a full height dense brand of around 40 mm high and up to roughly 600 mm long topping out around 75 to 85 m/min on one head. For anything else the mark decides the speed, so tell us mark height, character density and mark length and the answer comes back specific. Above the single head ceiling the specification is a second head sharing the mark on the same encoder.

Which source would our application call for?

The material decides it. Fibre at about 1064 nm on steel, stainless, aluminium, titanium and coated metal. MOPA where pulse duration, repetition rate and pulse energy have to move independently, typically 4 to 200 ns at 1.6 kHz to 1 MHz. Green at 515 or 532 nm on copper, brass, gold and silver. UV at 355 nm on glass, quartz, ceramics, PCBs and heat sensitive plastics. CO2 at 9.3, 10.2 or 10.6 micron on timber, board, acrylic and coated surfaces, where 9.3 micron into PET is the difference between a readable code and no code at all. Ultrafast where a heat affected zone is itself the defect. We specify and integrate the source the material dictates, and the parameters are established on your own material in the capability study.

Will it mark our material?

Send a piece and the capability study answers it with a measurement: the substrate and surface condition as received, the parameter set, the verification grade against the applicable standard, the durability result against the conditions you name, and the determined process window. The sample comes back with the settings and the margin the parameter set carries, and that number is the one written into the acceptance criteria.

How is a symbol grade committed?

A grade is a measurement, taken on your substrate with the reader class your operation carries. It is established in the capability study, then written into the acceptance criteria as a threshold with the standard and the test that closes it beside it. A typical criterion reads: Data Matrix legible after 90 degree caustic wash, verified per ISO/IEC 15415, accepted at grade C or better after 30 cycles.

How are MIL-STD-130 and AS9132 handled on our parts?

Light Lane specifies to and verifies against MIL-STD-130 and AS9132 where your operation works to them, and supplies the documentation for handover. They are real published standards you can buy and read, and on a programme they appear inside acceptance criteria as thresholds with test references rather than in a badge row. The same discipline applies to ISO/IEC 16022, ISO/IEC 15415 and ISO/IEC 15416.

What sets Class 4 laser safety on an installed line?

Enclosure, interlocks, beam containment and a rated safety circuit, designed into the install against ISO 11553-1, IEC 60825-1, IEC 60204-1 and ISO 13849-1 rather than added to it afterwards. The safety class belongs to the installed arrangement, so it is stated for your installation once the source and the enclosure are fixed, and it is part of the install specification rather than an accessory.

What happens if the physics routes our job to a coating, a plate or a label?

The numbered capability report says so and states the reason, so a buying group holds an engineering result before capital is committed. The method selection table above sets out the cases where that happens and what we specify in each one, including the laser reactive coating route that removes the substrate question from the line entirely.

How much does the software update change what our line produces?

A golden fixture suite holds emitted machine output stable across three dialects, 42 fixtures across two images and seven processing variants, with a conformance validator checking output against the dialect rules, so a change to the geometry engine shows up as a failed fixture in continuous integration rather than as a different part on the floor. Marking parameters and machine profiles for every material and mark type are documented and resident on your site, so a setting is recovered rather than rediscovered.

Send us a part

Post the part in the material and surface condition it actually arrives in, and tell us what happens to it afterwards: the wash, the chemistry, the abrasion, the years outdoors, the scanner class your operation carries, and how fast the line moves. You get back a numbered Marking Process Capability Study stating the substrate and surface condition as received, the parameter set, the verification grade against the applicable standard, the durability result against those conditions, and the determined process window. That is the shortest route from a published envelope to a number that belongs to your own material. Where the line itself is the question, book a site scope and the survey covers the doorways, the power, the extraction and the vibration alongside the part.

Send us a part Book a site scope

Last updated August 21, 2026