Engineered once, proved on the hardest line, replicated across the group

Integration at group scale

A record-keeping problem is never allowed to become a production problem

Every brochure in the category describes a system that works. This is what the station does when something upstream moves. Cache depth and hold duration are sized against your run length and line rate at the specification stage, and each behaviour below is written into the interface specification before build.

Integrating a marking line requires no inbound path into the plant

One posture, reviewed once by your group security function and deployed unchanged at every site. A mixed estate is normal: a plant sitting on the corporate network and a plant sitting on an isolated OT network behind a full air gap run the same build, and at the air-gapped site the licence lifecycle runs locally with it.

• Outbound only. The station initiates and your endpoint answers, so a site joining the programme opens no port for marking and your ERP stays behind whatever the group already puts in front of it.
• Least privilege, issued and revocable by you. Read access to the job or batch fields the mark is composed from, write access to the single field carrying mark confirmation, and a PLC signal tap. Credentials are issued per site and revoked per site, so withdrawing one plant's access is a single action that touches nothing else in the group. No personal staff login is requested or wanted.
• Read-only PLC posture. The station reads the agreed tags and leaves ladder logic and function blocks untouched. Any write is scoped as a specific handshake and signed off separately. OPC UA runs certificate-based security where the environment provides it. Each site's tag list is signed off by that site's control engineer and recorded in the reference design, so what a station reads is auditable from the document rather than from the controller.
• A site with no route out runs the whole licence lifecycle on its own. The application mints a Device Setup Code carrying the device identifier, an Ed25519 public key, the device name and the platform. The portal returns an activation package containing an activation JWT and a lease JWT. The application verifies the RS256 signature, the issuer, the device identifier and the public key hash entirely locally. Offline lease refresh and offline deactivation both use Ed25519-signed requests, and the deactivation code persists across a restart so the device slot is provably released.
• Licence integrity controls. State becomes expired if the system clock is set more than an hour before lease issuance, and local activation is cleared the moment the server reports the device unbound.
• Secrets held in the operating system credential store: macOS Keychain, Windows Credential Manager, freedesktop Secret Service. Never a flat file on disk.
• Supply chain integrity. Updates are fetched from a signed version manifest, and every download is verified against a published SHA-256 checksum before installation.
• No retention of your production data. The station holds what it needs to complete the parts in front of it. Adding sites therefore adds no second production data store for the group to govern, retain and produce on request.
• The AI assistance in the desktop application is the one part that needs a connection. On an isolated site it is switched off and everything else runs unchanged.

What a programme covers, end to end

Each of these is scoped and priced as its own line rather than folded into a single number labelled 'system', so a buying group can see what buys what.

• Scope. An on-site survey of the actual parts, the actual line and the actual conditions: material and surface state at the marking point, part presentation, line rate, floor space, power, extraction, and what the mark has to survive downstream through wash-down, weather, heat, handling and coating. The pilot site is surveyed in full and every following site is surveyed against the reference design, so a difference in material, presentation or line rate is recorded as a difference rather than assumed away.
• Source selection. The source follows the material: fibre and MOPA on steel, stainless, aluminium and coated metal, green on copper, brass and precious metal, UV on glass, ceramics and heat sensitive plastics, CO2 on timber, board and coated surfaces. On a group programme each material family is assigned one source and one parameter set inside the standard, so the second site running that material inherits a setting rather than deriving one. The wavelength selection matrix, and the conditions attached to every figure in it, are on the engineering specification.
• Engineering. Marking process design and capability trials on your own material, fixture and part presentation design, safety and extraction layout, and cycle time proved against your line rate. A typical timber brand on a single galvo head holds across 40 to 120 m/min. 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. The five variables that move that ceiling are mark height, character density, total mark length, line speed and available power, and beam time is path length, not area. Published figures are typical envelopes from our own testing, and the binding number is the one written into the acceptance criteria after the capability study on your material.
• Build. The core is deployment-form independent: an inline galvo station marking a moving line, a marking head added into a fibre or CO2 cutter the plant already owns so the cut pauses, the code lands in the fixture and the cut resumes, a counterbalanced arm, a handheld rig for assets too large to come to a station, and the desktop application on a single machine. The rigs are designed and built in our Nelson facility: conveyors, rotaries, fixturing and jigs, extraction and filtration, safety enclosures and interlocks, camera and vision alignment, and counterbalanced arm and handheld heads. Class 4 safety is enclosure, interlocks, beam containment and a rated safety circuit to IEC 60825-1 and ISO 11553-1, specified with the station rather than added to it.
• Integration. A versioned written interface specification against your ERP or MES, built and tested on representative production data before it goes near the live system, with a named owner on each side and the failure behaviour above written into it clause by clause.
• Commissioning. Install, alignment, calibration, safety sign-off, integration cutover under your change control, and a first-run validation pass on real parts at real line rate against acceptance criteria agreed in writing before the build started.
• Training. Operators, shift leads and maintenance trained on the installed system using your own jobs, until they run it without us in the room. Cover is trained at each site: the named owner and at least one backup on every shift pattern, working from the course the pilot site ran.
• Support. A support tier scoped to the operation, with severity tiers defined in plant language, a named escalation ladder, remote diagnostics on the software and integration side, on-site attendance where the tier calls for it, and software updates for the life of the system. The structure is published below and the committed times are written into the support agreement before commissioning.
• Spares. A holding agreed at commissioning against the parts that would actually stop the line, sized to the programme and held on site, rather than ordered from overseas after the first failure. The holding is sized per site rather than per group, because a spare sitting at another plant is not a spare.
• Documentation. As-built documentation with equipment make, model, serial and distributor; marking parameters and machine profiles for every material and mark type, saved on the system and written down so a setting can be recovered rather than rediscovered; fixture and rig drawings; the interface specification as built; operator procedures; the maintenance schedule with stated intervals; the recommended spares list with lead times; the written support arrangement; and warranty documentation.

Reference design, then controlled replication

A group standard that was never proved on a line is a document. The order runs the other way: one line, accepted on its own numbers, then replicated under change control.

1. Step 1

   Prove it on the hardest line

   The first site is chosen on difficulty rather than convenience: the worst material, the busiest shift, the oldest ERP instance, the most awkward part presentation. If the marking process holds there, it holds at the easier sites, and the parameters carry across instead of being rediscovered.

   Why it matters: A pilot chosen for an easy win tells a group nothing about the sites that follow.

2. Step 2

   Accept it against criteria written before the build

   The pilot is accepted against criteria agreed in writing before anything was built: mark permanence on your material under the conditions you state, read rate at line speed on the scanner class your people carry, verification grade per ISO/IEC 15415 at the threshold your customer requires, cycle time against the real line rate, and the round trip to your ERP confirmed end to end with the completion record landing against the right job.

   Why it matters: The pilot's acceptance record is the number every following site gets measured against, so it has to exist before the first build starts.

3. Step 3

   Turn the signed-off line into a reference design

   Once accepted, the line becomes a documented reference design: 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 next site starts from that document rather than a blank page, and any change to it is a change to the group standard.

   Why it matters: A reference design is what turns the third site into a deployment instead of another project.

4. Step 4

   Split the group standard from what each site sets locally

   Every site has its own building, power supply, extraction route, shift pattern and scanner fleet, and its own variation in the parts. Before rollout, the reference design is split into what stays fixed across the group, marking parameters, the interface specification, identifier standards, safety design and support tier, and what each site sets for itself, line rate, part presentation, power, extraction routing and scanner fleet. A plant manager can see which decisions are theirs and head office can see what stays standard.

   Why it matters: A rollout that assumes every site is identical stalls at the first one that is not.

5. Step 5

   Replicate site by site, with the same sign-off each time

   Each following site 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. Group standard, group documentation, group support tier, local proof. Sites go live on their own schedule rather than waiting for the whole programme to finish.

   Why it matters: A site holding its own signed acceptance record stops treating the rollout as something done to it.

How the programme is staged

A programme this size is rarely approved in one signature. Each stage is scoped, quoted and accepted on its own, and each leaves 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 three programme risks we design out before build

Marking programmes are decided by three things that have nothing to do with the laser, and all three are visible early enough to design out. Each one is a scope item before the build starts rather than a discovery in month four.

Support is committed by severity tier

The tier definitions are written in plant language. What is handled remotely, what triggers site attendance and the escalation ladder are fixed across the group. The committed times are set per programme against your site location, shift pattern, on-site spares holding and whether the line can run degraded, priced in the proposal you approve, and written into the support agreement before commissioning.

Questions asked before the first stage is approved

Capital structure, staged payment, prequalification and the acceptance document set are on programme assurance. Materials, mark types, speed envelopes and symbol grades are on the engineering specification.

One control layer, different source, different substrate, different line rate

The same composition engine, the same connection patterns with their authentication models and failure edges, the same acceptance document set and the same handover pack are specified across every substrate a group runs, and only the source and the parameter set change with the material. That is what lets a group holding a timber site, a meat plant and an engineering shop run one standard rather than three separate programmes. The shift is substrate blind: a retailer moving to 2D scanning does not care whether the object is a chilled carton or a returnable crate, so any marking approach that works on one substrate alone is disqualified from being the answer by definition.