Hope for solving the single-use packaging crisis comes in the form of a collaborative technological effort between Zebra Technologies, Polytag and OEM Automatic. MEPCA talks with Tim Stoddard of Zebra and Jon Anderson at Polytag about this innovation and the partnership that shaped it.
“I think sustainability and innovation are quite intrinsically linked,” Tim Stoddard, General Manager EMA, Zebra Technologies.
A principal challenge in achieving circular economy in packaging and improving recycling rates is how to accurately track single-use plastics.
With new Extended Producer Responsibility (EPR) regulations now in force, which financially penalise businesses for single-use packaging, it has become ever more important for companies to be able to provide accurate data about where and when their packaging is recycled or disposed of.
The problem is that current recycling sorting technology is simply not able to deliver the level of accuracy required.
Fortunately, a collaboration with Polytag and Zebra Technologies has delivered an ingenious solution by developing a system that combines invisible UV tags, infra-red and machine vision to trace individual items of recycling from the producer to recycling centre. The system can be retrofitted in recycling centres, and it provides accurate, real-time data to businesses.
Driving the project was Polytag, a leading provider of innovative QR code and UV tag & trace solutions for circular economies, and based in Wales, the second-best country for recycling in the world. Speaking at the First Friday Club Editors’ Briefing, Jon Anderson, Chief Technology Officer, Polytag, detailed how the solution works and where Zebra’s machine vision expertise comes in.
As a method of identifying the product, they applied UV ink to the packaging at the point the label is manufactured. It cannot be seen by the consumer but fully surrounds the packaging and includes the GTIN and barcode data, in a universally readable format, to GSTL standards.
The next step was to develop the detection technology. For this they combined hardware from OEM Automatic UK, industrial automation experts and the third technology partner in this innovative collaboration, and worked closely with Zebra Technologies to leverage its Aurora Vision Library, powerful machine vision software, to achieve 100% accuracy in detecting and tracking waste packaging.
Speaking with MEPCA’s editor, Jon recalled that after the founding of Polytag and the first phases of research and development, they were seeking a more industrial solution for the project. They had been working in partnership with AMRC (Advanced Manufacturing Research Centre) Wales, which connected them with OEM Automatic UK, an exclusive supplier of Basler cameras and vision components, and later, Zebra.
“We explored quite a few options,” he explained, “of both technologies that were built within camera and also technologies that were more software based, and after trying a few we found that for this specific use case, although not developed for this type of detection, we found the Zebra algorithm very, very, very powerful and it allowed us to have a lot of control and flexibility with the work we did.”
Elaborating on what technologies were trialled prior to partnering with Zebra, Jon said, “Our system was using a hardware-based controller camera system, so the cameras themselves decoded directly inside there, and we determined that, when we looked at options, you want a more software-based approach.” They therefore looked at open-source software, and trialled and tested machine learning tools. But Zebra provided them with the way forwards.
As the partnership took shape, a key area Polytag and Zebra worked closely together on was the optimisation of the software and training of staff, combining their expertise.
When asked about the foremost challenges of the implementation for the proof of concept, it proved to be what separated the project from the more typical industrial application of a decoding system: an uncontrolled environment.
“We’re dealing with materials that are running at different speeds, of different qualities, of different levels of damage and elements […] You start to see performance issues where you’re trying to code without really specific boundaries. So, where in a more industrial application you can refine the cameras and the lighting to be very specifically targeting areas which have minimal image size. We needed in some of these systems to have four cameras across a belt of 2.8 meters wide, but still be able to decode these at speed.”
Zebra worked with Polytag to help iterate and refine the software as they faced these and other challenges.
For Zebra, a project such as this had a great deal of synergy with both internal sustainability ambitions and their wider commitments to help customers with circular economy projects. As Tim Stoddard explained, sustainability is something that both he and the wider company have long championed.
“So, personally, I have been sponsoring a number of our green earth initiatives. We have a number of internal and external practices. We have sustainability objectives.” These include the Zebra Emissions Targets (Scope 1-3). Both to be achieved by 2030 against a 2020 baseline, Zebra is aiming for a 50% absolute carbon emissions reduction at Zebra sites (i.e., scope 1 and 2) and 15% absolute carbon emissions reduction in supplier manufacturing and customer product use (i.e., scope 3).
As well as these internal goals, Zebra has the “Green Herd” a team of volunteers working to differentiate the brand in terms of sustainability, with regional, out-of-hours initiatives, such as tree planting.
“I would say I’m passionate about that,” he continued. “I think, you know, as a company, we talk about innovation as being one of the pillars of our culture. I think sustainability and innovation are quite intrinsically linked.”
For these reasons and more, the collaboration with Polytag was an ideal project for Zebra.
Since completion of the trial, the system is gaining momentum, with large consumer brands such as Ocado utilising Polytag for traceability.
Jon: “What we do have now is a large network throughout the UK where we are again collecting data every single day and building a really interesting picture of our customers product as it’s getting recycled.” And while he couldn’t share the details explicitly, he said, “We’ve got a number of interesting projects going on further afield across Europe and into Asia, but nothing quite crystallised to this point yet.”
Tim is confident in the extensive applicability of the underlying technology: “It’s pretty horizontal in terms of different use cases. Traditional use cases are typically quality inspection, quality assurance, quality control. I would still say that’s a generic or one of the primary use cases. We’ve seen this expansion requirement around damage detection. We’ve seen some interesting use cases around batch code visibility. Ultimately, asset tags that are not visible to the human eye and be able to pick those up. And we have one customer move from, I think it was about sort of a 35 to 50% accuracy using human eye to 90%.
“It’s not only a time saving, but also an accuracy saving … the learning capabilities are accelerating rapidly. For us, it’s about understanding the customer problem and then leveraging the solution to be able to address that. And those are probably the typical use cases we see.”
Did they foresee there being other future collaborations between them outside of this project?
“Yeah, absolutely.” Jon said, “Zebra would be our first point of call if we needed any enhancement. For a lot of this, expect our technology is going to continue to broaden its existing. I can’t see us inventing something very different away from our current key focuses. But we continue to work with Zebra as we need support across the different parts.”
Tim: “And let me just compliment that. Two of the themes that are forefront of mind for me are security and sustainability. And certainly, from a sustainability perspective, we’re looking at how we can address customers’ problem statements, whether it be optimising workflows, whether by minimising errors or helping them achieve their sustainability objectives. So, I think, you know, through the partnership, we’ve got great relationships and I’m sure we’ll identify some additional problem statements to go address together. And we’re passionate about doing that.”
For Zebra, this has been a prime example of what they term their “ecosystem”, a key differentiator for them as a brand, which Tim explained at a high level:
“So, we bring solutions to market and our ecosystem will add and create value around it. And that’s exactly what OEM Automatic and Polytag have done here. So, we have this overall ecosystem, which is whether it could be resell partners, it could be influence partners or independent software vendors.
“We [also] have something very specific, which is our advanced machine vision specialist program, which is designed to reward those specialists in those spaces who are willing to invest in it. And this three-way partnership is a perfect example of that; of where we can work with the right ecosystems to be able to address the right customer problems in the most effective manner.”
Tim has previously described the company as a chameleon; “as they are a brand that hides in plain sight.” The analogy perfectly captures how common place Zebra’s technology is, intersecting with so many aspects of peoples’ lives, but rarely, if ever, noticed. This project serves as another fine example.
