Traceability Weekly™ | Product Identity Is Becoming a Continuous Information Loop | 18 September 2026

There is a noticeable shift taking place across product identification and traceability.

For years, most systems were built around individual events. A printer created a code. A camera checked it. A scanner read it. A pallet was labelled. Each stage did its job and passed the product downstream.

The strongest developments this week suggest that model is becoming less useful. Identification, inspection and product movement are increasingly being treated as parts of one continuous information process rather than isolated operations on the line.

That is visible in four very different developments: P&G and Siemens scaling AI inspection across global manufacturing, Domino completing its own rollout of GS1-powered QR codes, Avery Dennison launching active Bluetooth labels, and Scott Technology introducing a compact palletising platform aimed squarely at constrained brownfield environments.

The technologies are different, but the direction is consistent.

Mark → Label → Verify → Inspect → Package

is increasingly being supported by a second layer underneath it:

Identify → Verify → Monitor → Record → Act

P&G turns AI vision into repeatable factory infrastructure

The most significant development this week came from P&G and Siemens.

P&G is expanding an AI-based visual inspection system across its manufacturing operations worldwide, with Siemens reporting scrap reductions of between 10% and 20%, depending on the product. New installations can also be commissioned five to ten times faster than more traditional bespoke machine-vision systems.

Those figures matter because the bigger challenge in industrial vision has never simply been whether a camera can detect a defect. The harder problem is whether an inspection system can be deployed repeatedly across different lines and plants without effectively having to engineer every application from scratch.

P&G’s manufacturing environment is a good example of why that is difficult. High-speed consumer products often involve materials that wrinkle, stretch, overlap or vary naturally in appearance. Conventional rule-based systems can struggle when “good” product itself has a degree of visual inconsistency.

The AI-based approach is designed to cope with more of that natural variation while still operating at production speed. Inspection results are processed close to the equipment using Siemens Industrial Edge and can be fed directly back into line control.

That turns machine vision into something more than a camera making a pass/fail decision. It becomes part of the production-control loop.

The more interesting number may therefore be the claimed improvement in commissioning speed rather than the scrap saving. If a manufacturer can take a successful inspection architecture and reproduce it across multiple plants without rebuilding the application every time, the commercial case for large-scale machine vision becomes much stronger.

For vision OEMs and integrators, that also changes the question customers are asking. It is no longer simply, “Can you inspect this product?” Increasingly, it is, “Can this architecture be deployed across 20 lines and five sites without becoming 20 separate engineering projects?”

That requires a blend of machine vision, AI, edge computing, PLC integration and process knowledge.

Domino’s GS1 rollout shows the difficult part sits behind the code

Domino also produced one of the more useful GS1 implementation stories this week.

The company has now completed the rollout of QR Codes powered by GS1 across its own newly manufactured equipment and consumables, allowing users to access information such as technical specifications, certificates, safety data and reordering information through a single standards-based code.

The more valuable part of Domino’s announcement, though, is not the QR code itself.

The company makes clear that implementation required product data to be reviewed and standardised, systems to be updated, processes to change, teams to be trained and long-term data governance to be addressed.

That is a useful reality check for manufacturers approaching GS1 2D as though it were simply a packaging-artwork exercise.

The physical code is the visible endpoint. Behind it sits the product identity, the master data, the systems that manage that information, the resolver or digital destination, the printing process and the governance that keeps the whole structure accurate over time.

That is why GS1 Sunrise 2027 should not be thought of simply as a transition from a 1D barcode to a QR code or Data Matrix.

Sunrise 2027 remains an industry transition ambition rather than regulation, but the practical work behind it cuts across coding and marking, labelling, barcode verification, software, IT and packaging operations.

Domino’s own implementation makes that point particularly well because the company is effectively becoming its own case study.

The lesson is straightforward: printing the code is usually the easy part.

Avery Dennison pushes identification beyond the scan point

Avery Dennison’s new AD Visi range takes product identification in a different direction.

The company has launched printable Active Bluetooth Low Energy labels designed for time-sensitive and high-value supply chains. The initial products include AD Visi Way, for real-time location tracking, and AD Visi Secure, which adds tamper evidence and can associate a tamper event with time and location.

This matters because conventional traceability is still largely event-based.

A barcode tells you something when somebody scans it. Passive RFID tells you something when the tag passes through appropriate infrastructure. Between those events, the system may know relatively little.

An active BLE label changes that relationship. The object can broadcast rather than wait to be interrogated.

For higher-value goods, pharmaceuticals, returnable transport assets or sensitive shipments, that creates the possibility of knowing not only where something was last scanned but what happened between conventional checkpoints.

That does not mean BLE replaces QR, Data Matrix or RFID. The more likely direction is layered identity, with different technologies doing different jobs.

A standards-based 2D code may provide universal identity. RFID may handle high-volume automated reading. Active sensing may be justified where real-time location, condition or tamper status has enough commercial value.

The traceability architecture becomes richer as the value or risk attached to the product rises.

Scott Technology tackles one of EoL automation’s less glamorous constraints: floor space

At the other end of the line, Scott Technology launched its NexPAL 200HL palletising system this week.

The company says the system can process more than 35 cases per minute while using around 25% less floor space than comparable palletising systems. Four units are already scheduled for deployment at a major US poultry processor.

The footprint figure is important.

A lot of automation discussion assumes manufacturers have the luxury of designing new facilities around the technology. In reality, a large proportion of projects are brownfield. Plants want more throughput, more automation and less manual handling without having more floor space.

That makes compactness a commercial enabler rather than just a machine specification.

Scott’s design is aimed at precisely that problem, using a side-mounted rotating row pusher so case infeed and row formation can happen in parallel.

The broader relevance to traceability is that EoL equipment increasingly sits at the boundary between production and logistics.

A finished case may need to be identified, checked, conveyed, palletised, labelled, associated with a pallet ID, moved into storage and linked to an order or shipment.

That brings palletising into direct contact with print-and-apply systems, barcode readers, machine vision, PLCs, MES and WMS infrastructure.

The palletiser may be mechanical equipment, but the process around it is increasingly data-driven.

Regulatory & Standards Watch

The most immediate deadline remains the UK’s new Vaping Products Duty and Vaping Duty Stamps Scheme, which begins on 1 October 2026.

Digital duty stamps have been available since 1 September and are designed to support authentication and traceability through scanning. HMRC’s system also supports aggregation, allowing individually stamped products to be grouped beneath an aggregate code for movement through the supply chain.

For affected manufacturers, this is now less a regulatory-planning exercise and more a production-readiness test.

The critical questions are practical: can the stamp be applied correctly, can it be scanned at production speed, does aggregation work reliably, are the right records being generated, and can warehouse processes handle the resulting identifiers?

That is where regulation becomes an engineering problem.

GS1 2D is also worth watching this week, not because there is a new deadline, but because Domino’s own implementation provides a useful example of what “readiness” actually means.

The move is often illustrated as 1D → 2D. In reality it is closer to:

data → identifier → systems → print → verification → digital destination → governance

That distinction matters because it creates work well beyond scanners and printers.

PPWR should also remain on the horizon. The regulation has been generally applicable since 12 August, although many requirements phase in over time. The practical implication for manufacturers is that packaging, labelling and identification equipment selected today increasingly needs to be flexible enough to cope with changing materials, pack formats and information requirements tomorrow.

What the week tells us

The four main stories this week are not really about AI, QR codes, Bluetooth labels or palletisers in isolation.

They are about continuity.

P&G is making inspection capability repeatable across plants.

Domino is linking a physical code to maintained digital information.

Avery Dennison is extending visibility between conventional scan events.

Scott Technology is increasing the physical automation through which identified products move.

Taken together, the production line is becoming less dependent on isolated information events and more dependent on a continuous relationship between the product, its identity and the systems around it.

That is the deeper change behind:

Mark → Label → Verify → Inspect → Package

The individual technologies still matter.

But increasingly, the competitive advantage comes from making them behave like one connected production and traceability system.

Talent Signal

The same convergence is showing up in the skills market.

Scaling AI vision globally requires more than camera expertise. It pulls together optics, AI, edge computing, controls and process engineering.

GS1 2D implementation requires coding knowledge, product data, software, standards and governance.

Active intelligent labels combine materials, electronics, wireless systems and supply-chain software.

Modern EoL systems combine mechanical automation, controls, robotics, vision, identification and warehouse integration.

The market still needs deep specialists. But the people becoming particularly valuable are the ones who understand the interfaces between those specialisms.

Traceability Weekly™ is produced by RoboEdge Talent, covering the companies, technologies, regulation, standards and specialist talent shaping global Product Identification, Traceability & End-of-Line Automation.