Integration architecture
The connection patterns with their authentication models and failure edges, the data round trip, and the security posture an IT lead can forward.
Twelve steps, in the same order every time, arranged as a V. Feasibility runs before specification: your own parts are marked, graded and returned with a numbered Marking Process Capability Study before a requirement is written. Acceptance runs before build: the mark rate the station holds, the verification grade, the cycle time and the round trip to the system that owns your job are agreed in writing and mapped to the named test that closes each one. The Light Lane Control Layer is then built, run end to end against representative parts before it ships, and commissioned on your line at real line rate. Every step names the artefact it produces, and those artefacts are what the programme is accepted against.
Laser sources, galvo heads, enclosures and interlocks are catalogue equipment from established manufacturers, and any competent buyer can order one this afternoon. The engineering sits either side: putting a permanent, readable mark into a part that is wet, hot, oily, coated or moving, presenting that part to the beam the same way on every shift, drawing the whole mark inside the window the part spends in the field of view, and making the code on the part the identifier your ERP or MES already generated for that job, with the completion written straight back against it.
Each step below closes one specific failure mode, by design rather than by vigilance. Material behaviour is characterised on the actual part before a number is written down. The fixture is designed from part presentation that was measured and photographed on your floor. The interface is a versioned document with a named owner on each side and its failure behaviour written in clause by clause. The operators meet the system on their own jobs, and a named operator owner and a named backup are agreed with the plant manager before commissioning.
The programme date is set by four drivers rather than a house average: hardware lead time on the source and the head, your own change control calendar, the number of materials and mark types the capability study has to characterise, and whether the interface pattern is already sanctioned in your environment. A single station on a characterised material with no integration sits at the short end. A custom rig on a new substrate with an interface that has to pass a change board sits at the long end, and most of that time is lead time and your own approvals rather than build. The dated programme is issued with the quote, against those four drivers, and the gates it is measured at are fixed before the first stage is approved.
A marking programme is written as a V. Specification descends the left arm, verification climbs the right, and the Requirements Traceability Matrix runs across the middle carrying every requirement over to the test that closes it. The pairs below are what decide whether a marking line is still trusted a year later. The full document set, the acceptance gates and the payment structure behind them are on programme assurance.
| Specified on the way down | The named test that closes it, on the way up | Where the result is recorded |
|---|---|---|
| Mark permanence on your substrate, in the surface condition the part actually arrives in at the marking point | Trials on your own parts at step three: the source the material dictates, the parameter set established, the mark verified, and durability run against the conditions you state | The numbered Marking Process Capability Study, carried into the Marking Requirements Specification |
| Symbol quality: Data Matrix generated to ISO/IEC 16022, graded per ISO/IEC 15415, linear symbologies per ISO/IEC 15416, accepted at the grade your scanner class supports | Verification in line at rate, on the scanner class your operation carries, at the accept threshold written into the criteria before the build | The Factory Acceptance Report, then the Site Acceptance Record signed on your line |
| Cycle time against the real line rate at the marking point, and the mark rate the station holds | Your parts run at your cycle time on the staged system, then again at commissioning on the line itself. A typical timber brand holds across 40 to 120 m/min on a single galvo head, and a full height dense brand tops out around 75 to 85 m/min, 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 | The Factory Acceptance Report, then the Site Acceptance Record |
| The data round trip: which fields cross the boundary in each direction, over what transport, under what authentication model, and which single field carries mark confirmation back | The integration built and exercised against representative production data before it goes near the live system, then cut over under your change control with a documented rollback and the completion record confirmed landing against the right job | The Functional and Interface Specification as built, versioned and carrying a changelog |
| Failure behaviour when a system underneath the Control Layer moves: a scheduled ERP maintenance window, a network drop mid-shift, a signal conflict at the head, a schema or credential change | Each behaviour exercised deliberately at factory acceptance: the pull-ahead cache carries the station through the window, completion records queue and replay in order carrying the original mark timestamp rather than the write time, and the station holds rather than composing from stale data | Clause by clause in the Functional and Interface Specification, evidenced in the Factory Acceptance Report |
| What the station does with a part that fails to mark or fails verification in line, hold or divert, chosen by you, and what the operator sees when it happens | Run on the line during commissioning with the people who will hold the shift, on your own jobs, until they run it without us in the room | The operator procedure in the Operational Handover Pack, signed off at the Site Acceptance Record |
| Class 4 laser safety: enclosure, interlocks, beam containment and a rated safety circuit, specified to ISO 11553-1, IEC 60825-1, IEC 60204-1 and ISO 13849-1, with the extraction route designed against the fume the material and its treatment chemistry actually produce | Safety and extraction sign-off at install, with the interlock configuration verified as installed and the safety circuit proved before the first production shift | Safety documentation, the extraction layout and the interlock configuration in the Operational Handover Pack |
Steps one to six descend the left arm of the V and produce the documents the system is built against. Step seven is the build at the base. Steps eight to twelve climb the right arm, closing each requirement against a named test. The detail changes with the job. The order does not. Each step states what happens and what it needs from you before it can start.
1. Qualification call: the part, the code, and the system that already owns it
Thirty to forty-five minutes, usually a video call. What the part is, what it is made of, how many run in a shift, what the mark has to carry, which system generates that data today, and what the mark has to survive downstream: wash-down, weather, paint, heat, abrasion, years outdoors, and the scanner class carried at the far end of the chain. We also ask what is driving the change, because an audit finding, a customer mandate and a recall each set a different date. Where the physics routes the job to a label, a plate or a tag instead, the call says so and states the reason. From you: a photo of the part and a rough volume.
Why it matters: Forty minutes with the part described properly selects the method. Nearly every failed marking project was decided in a meeting room with no part in the room.
2. Site scope on your floor
Half a day at the plant on the actual parts, the actual line and the actual conditions: floor space, power supply, extraction route, where the part sits in the flow and whether it can pause there, how it is oriented and presented, what it is coated with, how hot, wet or oily it is at that exact point, the line rate the station has to hold, and who handles the part next and with what scanner. Part presentation is measured and photographed, because it decides the fixture and the fixture decides the rest. Alongside it, integration discovery: which system already owns the job data, where identity breaks today, and what your control engineer will sanction on the network. The scope ends with a decision on feasibility, in writing. From you: site access, an hour of the production supervisor's time, and permission to take photographs.
Why it matters: A code nobody can scan is the same as no code, and the reason is usually standing on the floor rather than sitting in the software.
3. Marking Process Capability Study on your own material
You send real parts off your real line, in the condition they are in at the point the mark would be applied, including the worst ones. The trials establish the source and wavelength the material dictates, then power, speed, frequency, pulse regime and mark geometry, and verify the result the way your operation will verify it: read rate on the scanner class your people carry, contrast and verification grade against the standard your customer requires, and durability against the conditions you named. The parts come back with a numbered report stating the substrate and surface condition as received, the parameter set, the verification grade against the applicable standard, the durability result and the determined process window. From you: a box of representative parts, including the worst ones.
Why it matters: Marking is a materials process. Anodised aluminium, mill-scaled steel, green timber, powder coat and food-grade polymer each behave differently under the same beam, and surface moisture on green timber absorbs energy while escaping steam disturbs the mark as it forms. This step runs before a requirement is written, not after, which is what converts every published envelope on this site into a figure measured on your own part.
4. Marking Requirements Specification, agreed before anything is built
Both sides write down what the finished system holds: mark permanence on the material characterised at step three, read rate at line speed on the scanners your people actually carry, verification grade per ISO/IEC 15415 for 2D and ISO/IEC 15416 for linear at the threshold your customer requires, cycle time against the real line rate, what the station does with a part that fails to mark, what happens when the network drops mid-shift, and the round trip to your ERP or MES confirmed end to end. Every requirement is carried into the Requirements Traceability Matrix and mapped to the named test that closes it, with the source, the verification method, the test reference and the acceptance threshold in the same row. From you: sign-off from whoever owns quality or compliance, since they are the party who accepts the finished system.
Why it matters: Acceptance criteria agreed after commissioning are a negotiation. Written before the build, they are engineering, and they are why a factory acceptance produces a pass or a fail rather than an opinion.
5. System and interface design, written, signed and versioned
The system specification covers the marking process and its parameters, the source and head selected against the material, the fixture and part presentation, safety and extraction, and the operator workflow at the station, with cycle time proved against your line rate. Alongside it, the Functional and Interface Specification states how the Light Lane Control Layer and the system that owns your job exchange data: fields, direction, format, transport, authentication model, test data, deployment and rollback, and what each side does when the other is unavailable. The connection pattern is chosen with your IT team against what your environment already runs, from REST or SOAP, a read-only database view, a watched file drop, a message queue, an OPC UA or Modbus TCP tag read, or middleware your organisation already operates. On SAP the path is whatever your Basis team sanctions, usually existing middleware, and approval takes longer than the build. From you: a named technical contact on the systems side and a sample of real job data. Both documents are signed before the build starts.
Why it matters: An interface that lives in two people's heads stalls the moment either of them moves on. Written and versioned with a changelog, a line commissioned this year and a line commissioned next year run the same interface at a known version.
6. Itemised quote, staged against the acceptance gates
Machine, marking head, fixture and rig, software, integration, install, commissioning, training days, documentation, spares holding and support tier appear as separate lines with separate prices rather than folded into a single line labelled system. A larger programme is staged: capability study and scope, then design and specification, then build and pilot, then controlled replication, each scoped, priced and approved on its own, and each leaving the plant holding something usable whether or not the next stage is approved. Payment is released against the gates, so exposure at any point is capped at the value of the stage in progress and the final release is tied to the Site Acceptance Record signed on your line. The gates, the capital structure and the prequalification pack are set out on programme assurance. From you: the approval path, and who signs at each stage.
Why it matters: Procurement approves a number it can break down, and a plant that has proved the first stage decides the next one with the result in front of it.
7. Build and configuration, with the rig designed and built in Nelson
Sources, galvo heads, enclosures and interlocks are ordered as catalogue equipment from established manufacturers through the distributor channel that fits the job. Everything around them is engineered here: fixtures and jigs, conveyors and rotaries, extraction and filtration, safety enclosures and interlocks, camera and vision alignment, counterbalanced arms for parts too large to come to a station, handheld rigs for yard and field work, inline galvo stations marking on the fly against the encoder, and a marking head added into a fibre or CO2 cutter you already own, so the machine pauses part way through a cut it is already running, the head fires the traceability code onto the part exactly where it sits in the fixture, and the cutter resumes the same job. The Control Layer is configured to your part library, materials and machine, and the integration is built against the signed specification and tested on your sample data.
Why it matters: The part is anonymous from the second it drops out of the nest. Marking it in the fixture means the part number never leaves the fixture, so it can never be attached to another part's record.
8. Factory Acceptance Protocol, run and signed before it ships
The complete system is assembled and run end to end at our facility: rig, fixturing, enclosure, extraction, safety circuit and software configuration, with the integration pointed at a test instance of your system where one can be provided and exercised against representative production data. Your parts go through it at your cycle time against the criteria written at step four, and the failure behaviours are exercised deliberately rather than assumed: the maintenance window, the dropped network, the missing or conflicting signal at the head, the schema change underneath the integration. A system that has not passed its Factory Acceptance Protocol does not ship, and you are welcome to witness it. From you: a second batch of parts, and a test endpoint where your IT team can provide one.
Why it matters: Install day is the wrong time to find out a fixture does not clear the part. Debugging on your site costs you production. Debugging in Nelson costs a day.
9. Install and commissioning at real line rate
Physical install, alignment and calibration, extraction and safety sign-off, software onto the operator machines, machine profiles tuned to the delivered hardware, and the integration cut over to your production system under your change control with a documented rollback. Then the first-run validation pass: real parts, real line rate, real operators, marks verified for read rate and grade on your own scanners, and the round trip confirmed end to end with the completion record landing against the right job, carrying timestamp, operator, machine and station. Commissioning closes with the Site Acceptance Record: what was tested against the criteria from step four, what passed, and any open item with a named owner and a date. From you: a planned install window, an electrician where the power supply needs work, and access to the line.
Why it matters: A system that has not run at line rate on real parts has been delivered rather than commissioned.
10. Operator training on your own jobs
We train the people who will run it, on the system as installed, using your work rather than demo files: normal operation, changing between parts and materials, what a good mark looks like and how to check one, what to do when a code fails verification, daily and weekly maintenance, consumables, and the first checks before the escalation ladder is engaged. Shift leads and maintenance are trained alongside the operators, and a named operator owner and a named backup are agreed with the plant manager as part of the scope. Training closes when your team runs the station without us in the room, and more than one person is trained. From you: the operators, released from the line for the agreed training days.
Why it matters: Naming the operator owner is the single strongest predictor we have of whether a line is still marking a year later, and a lone trained operator is a single point of failure who also takes annual leave.
11. Operational Handover Pack
You receive a document set produced for your operation: as-built documentation with equipment make, model, serial and distributor, marking parameters and machine profiles for every material and mark type you run, fixture and rig drawings, the Functional and Interface Specification as built at a version number, the materials database exactly as configured, the operator procedure, the maintenance schedule with stated intervals, the consumables and spares list with lead times, safety documentation, the written support arrangement, warranty documentation and the Site Acceptance Record. It is written so your own engineer picks the system up from it, and so an auditor follows an identifier on a part back to the job that made it.
Why it matters: The second install should cost less than the first. If it does not, the pilot was not properly documented.
12. Support tier, spares and the life of the system
The support tier scoped in the proposal starts at commissioning, with severity defined in plant language: S1 is a stopped line with no workaround, S2 is a line running degraded on a workaround, S3 is a change with the line unaffected such as a new material or a new part family, and S4 is a question, a training refresh or a documentation request. What is handled remotely, what triggers site attendance and the named escalation ladder are fixed in the agreement, and the committed times are set against your site location, shift pattern, on-site spares holding and whether the line can run degraded, then written into the support agreement before commissioning. Software updates run for the life of the system, fetched from a signed version manifest and verified against a published SHA-256 checksum before installation. Spares are agreed at commissioning against the parts that would actually stop the line and held on site. Warranty is manufacturer-backed through the distributor and administered by us. When the product line changes, the re-scope starts from your documented reference design rather than a blank page.
Why it matters: A stopped line at 3am is production rather than a ticket, and it never happens at a convenient hour. The system has to still be running in year four, on parts nobody has drawn yet.
These people are identified at the qualification call and named in the specification, each against a defined artefact they sign or supply. A programme that names them early runs on a plan rather than on availability.
| Role | What they decide | Where they appear |
|---|---|---|
| Operations or production manager | The outcome, the line rate the station has to hold, where the mark sits in the flow, and what a completed handover looks like | Steps 1, 2, 9 and 10: qualification, site scope, the commissioning window and training |
| Engineering or technical lead | Part, material and surface condition at the marking point, fixture and part presentation, safety and extraction design | Steps 2, 3, 5 and 8: site scope, capability study, specification sign-off and factory acceptance |
| Control engineer | Which PLC tags the station reads, the network segment it sits on, and the sanctioned path from the floor to the business system | Steps 2 and 5: site scope and interface design, and again at cutover |
| IT, ERP or MES owner | Which fields cross the interface in each direction, the transport and authentication model, the test instance, and change control at cutover | Steps 5, 8 and 9: the interface specification, integration testing and cutover |
| Quality or compliance | Code standard, verification grade and accept threshold, audit evidence, and how long marking records are retained | Steps 4 and 9: acceptance criteria sign-off and the Site Acceptance Record |
| Procurement or finance | Staging, the itemised scope, the gates payment is released against, and the capital route | Step 6, and again before each later stage is approved |
The Light Lane Control Layer reads the PLC tag that identifies the part at the head, pulls the job, batch or serial from the system that already owns it, composes the mark for that individual part, fires it at line rate, and writes the completion back against the job with timestamp, operator, machine and station. It connects to the systems already running the plant, including NetSuite, Infor M3, Microsoft Dynamics 365, SAP, Cin7, Unleashed and Fishbowl, alongside the WMS, TMS, CMMS, calibration management and yard management systems beside them, through defined connection patterns that each carry a stated authentication model and a stated failure edge. Your system generates the job, the batch and the identifier. The plant never gets a second database to maintain.
Integration architecture
The connection patterns with their authentication models and failure edges, the data round trip, and the security posture an IT lead can forward.
Programme assurance
The six named documents, the requirements traceability matrix, the factory acceptance gate and the exposure cap.
Multi-site programme architecture
Reference design, controlled replication, the outbound-only security posture and integration at group scale.
Handover is a defined deliverable with a list attached, produced for your operation, resident on your site and yours. It is published here so a buying group knows what it holds at the end of the sequence before the first stage is approved.
The document set, the acceptance gates, the exposure cap and the qualification pack are on programme assurance. Materials, mark types, speed envelopes and symbol grades are on the engineering specification.
Four drivers: hardware lead time on the source and the head, your own change control calendar, the number of materials and mark types the capability study has to characterise, and whether the interface pattern is already sanctioned in your environment. A single station on a characterised material with no integration sits at the short end. A custom rig on a new substrate with an interface that has to pass a change board sits at the long end, and most of that time is lead time and your own approvals rather than build. The dated programme is issued with the quote against those four drivers, and it is measured at fixed gates.
Real parts in real condition for the capability study, including the worst ones. Site access and an hour of the production supervisor's time for the scope. Sign-off on the acceptance criteria from whoever owns quality. A named technical contact on the systems side, sample job data for testing, and a test endpoint where one can be provided. A planned install window, an electrician where the power supply needs work, and the operators released from the line for the agreed training days. It is listed step by step above so each one is scheduled rather than discovered during install week.
A numbered report stating the substrate and surface condition as received, the parameter set, the verification grade against the applicable standard, the durability result against the conditions you state, and the determined process window. Where the physics routes the job to a label, a plate or a tag instead, the report says so and states the reason, so a buying group holds an engineering result before capital is committed. That report is the first checkable object in the programme, and its numbers are what the Marking Requirements Specification is written from.
A marking head is added into a fibre or CO2 cutter the plant already owns, so the machine pauses part way through a cut it is already running, the head fires the traceability code onto the part exactly where it sits in the fixture, and the cutter resumes the same job. Light Lane marks with a machine it did not sell, inside a job it did not write. Feasibility turns on the controller, the cutting head clearance and whether the work envelope physically carries a second head, so it is assessed at the site scope rather than assumed. On the desktop side, the same controller-aware core emits a single dialect-neutral motion program into GRBL, Marlin, Smoothieware and Generic, with power scaling read off the controller's own registers at connect time.
The station runs from a pull-ahead cache rather than calling the business system per part, so the physical line keeps marking, and 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. 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. 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 and never increments a serial itself. Every one of those behaviours is written into the interface specification before the build and exercised at factory acceptance.
Sources, galvo heads, controllers, enclosures and interlocks are standard industrial equipment from established manufacturers through their New Zealand channel, ordered through whichever distributor fits the job rather than a single brand. The parts that decide whether the line works are engineered here: fixtures, jigs, conveyors, rotaries, extraction, safety enclosures, vision alignment, counterbalanced arms, handheld rigs, inline galvo stations and cutter retrofits are designed and built in our Nelson facility, and the Control Layer is written there. Continuity is engineered rather than asserted: every marking parameter, machine profile and interface specification is documented and resident on your site, and any competent integrator picks the system up from the handover pack.
The first line is chosen on difficulty rather than convenience, accepted against criteria written before the build, and then documented as a reference design covering marking parameters per material, fixture and rig drawings, the interface specification at a version number, safety and extraction layout, operator procedure, maintenance schedule and spares list. The group standard is split from what each site sets locally before rollout. Each following site still gets its own survey, its own trials where the material differs, its own commissioning pass on its own parts and its own signed acceptance record. The multi-site mechanics are set out on multi-site programme architecture.
Bring the part, the line rate the station has to hold, what the mark has to survive downstream, and the name of the system that issues your job numbers. The scope covers the floor: part presentation, power, extraction, where the mark sits in the flow, and the path your control engineer will sanction on the network. Send a part ahead of the visit and the numbered Marking Process Capability Study is on the table when we arrive, so the first conversation runs on a measurement taken from your own material.
Last updated August 21, 2026