Undefined: The Missed Potential of Digital Product Creation

By The Interline

Clothing Tech is a member of SPESA and will be exhibiting at Texprocess Americas May 17-19, 2022, in Atlanta, Georgia.

Digital product creation has become a cornerstone of fashion. From enabling brands to keep creating new styles during the pandemic, when physical samples were impossible to produce, to replacing product photography, more apparel businesses than ever are now building and visualising their garments and footwear in 3D environments.

But as critical as DPC solutions have become for these purposes, the assets they create can struggle to stand on their own. For a 3D asset to be shared with a manufacturing partner and to be considered producible, it has to be accompanied by a set of technical specifications, a bill of materials, and an associated 2D pattern. For a digital product to be confidently sold to a consumer, it has to be closely, manually aligned to the physical product. Wherever a digital garment goes, it relies on support from other solutions.

Or, to put it another way, as fashion looks to extend its use of 3D garment and footwear assets, the industry is increasingly finding that those garments can only represent the product with several degrees of abstraction – creating disconnects between the digital and the physical in a way that could be starting to undermine the real potential of digital product creation.

“For all the success of 3D in fashion, I believe the industry has undergone only a partial digitisation, and it’s still yet to address fundamental problems in how digital assets are, and aren’t, used in key communication stages” explains Bill Wilcox, Founder of disruptive new 3D CAD supplier Clothing Tech, who The Interline spoke to for this feature.

“Whether it’s between the brand and the customer, who’s making a subjective buying decision based on a range of different qualitative criteria, or between the creative designer, the technical designer, and the factory, there are gaps in communication where humans are required to interpret, translate, and adjust, and where the whole design, development, and production cycle falls back on the same pre-digital processes. Brands and their suppliers have replaced physical samples with virtual ones, but the industry is yet to settle on a way to really capture, define, and communicate fundamental truths about its products for in-house teams, supply chain partners, or consumers.”

That idea of definition is a critical one, because unlike other product-driven industries where the exacting details of any given product are set in stone early in its lifecycle, fashion operates more loosely over often-lengthy go-to-market timelines. After receiving a 3D asset from a brand customer, for example, a manufacturer will likely recreate the associated block pattern for achievability or efficiency reasons. And as an initial creative design undergoes the rigours of commercialisation, processes line reviews, costing and sourcing are often allowed to run in parallel and to only later be consolidated – with the effect that the nature of a product is constantly in flux up until the point a production order is placed, and sometimes even beyond.

While many of those concurrent processes do happen digitally today, in both formal systems and across informal channels like email and Slack, the task of reconciling them all is typically placed at the doorstep of PLM, where disconnected parts are pulled together to create a cohesive whole: the technical specifications, or tech pack. And while the tech pack is the output of a series of digital process, it is, by itself, not a digital product definition. Critically, the tech pack is also only assembled once the product has undergone its full cycle of design and development, but before it goes through the interpretative phase of production, making it perhaps better labelled as a product snapshot rather than a full definition.

“One of the primary reasons that it takes so long to design, develop, and produce even a relatively simple garment is that product definitions in apparel are still so fragmented,” Wilcox says. “Parts of the truth are stored in one location and parts are stored in another, and because of that lack of cohesion, and that absence of a single point of reference, the route to market is held up by changes, re-alignment, and sign-offs – which everyone knows often run right up to the last minute before starting production.”

“It seems long overdue,” Wilcox adds, “for the industry to take a step back and look at defining its products fully, digitally. Not just digital fabrics, or digital trims, that accompany a tech pack, but a digital asset that contains the complete pattern, along with all construction and assembly details – all defined in one single digital twin, which then serves as the source of truth for every task that follows, from pattern adjustments and grading, to sewing operations and virtual photography. 3D design and simulation tools, as they’re used today in apparel and footwear, have focused heavily on product visualisation, which is something they’re very successful at, but I believe 3D CAD should have a much more ambitious goal: to replace the tech pack as the single source of definition for a product.”

That vision is something that Clothing Tech has set out to realise with its Garment Digital Twin platform, translating core principles and best practices for product definition from engineering-led industries where products are defined in precise detail, digitally, at the earliest possible stage.

As an example, in the automotive industry the digital asset created at or near the start of a vehicle’s lifecycle will serve to define every component, from upholstery to engine, and it will remain relatively static throughout that vehicle’s lifecycle as the single source of truth for what parts that vehicle contains, and what operations are required to put it together. A direct analogue in the apparel industry would be for a 3D asset for a garment to contain not just materials and meshes, but details such as stitching, lining, pockets, fabric details, construction details, labels, ties, prints and much more – everything, in essence, that’s needed to turn the digital asset into an identical physical product.

This is, as Wilcox sees it, the fundamental difference between a virtual sample – which is the goal towards which most 3D adoption and innovation in fashion has been directed – and a true digital twin. While the former fulfils an extremely useful purpose, the latter, as a precise replica of the garment created in a way that alterations made to it cascade automatically into other areas, has the ability to transform the way the fashion industry thinks about not only digital product creation, but product creation in general.

“If I want to change a stitch, or add a pocket to a garment that’s reached the tech pack stage today, where do I make that change?” Wilcox asks. “I could go to the tech pack itself and mark that change up, but then my virtual sample doesn’t get updated. Instead I have to communicate that change to the 3D designer who makes it manually. Or I could make the change on the 3D simulation, but then I need to make an alteration to my 2D pattern, or have the factory I’m working with alter the pattern – because none of these things are connected, even though they all relate to the same product definition. This is something that PLM manages at the bill of materials level, or in a set of technical specifications, but if we can define products digitally from the outset, in 3D CAD, we can connect every stage of the product journey such that the alteration I make affects the seam allowance, the grading rules, the material yield, the assembly operations and much more. You change the core definition in one place, and it changes the usage of that asset and the nature of that product everywhere that definition is called.”

A question does arise, though. While 3D design and simulation tools are typically pitched as providing creative empowerment at the ideation and experimentation phase, and then have the added benefit of providing extra accuracy during technical development and production, rearchitecting the creative process to generate the full spectrum of product attributes during initial creation seems like it could place an additional burden on designers who are not accustomed to building out products in this way.

And that burden would come on top of additional requests already being made of designers, who are being asked to create to cost objectives and sustainability goals, and to meld a creative and a commercial mindset. Is the task of creating an early stage digital definition for every product too large to place on the creative community?

This is a question that Clothing Tech is looking to answer in two ways. First, through automation: the company has designed its Garment Digital Twin solution to automate the creation of 3D digital garments, extracting data from existing patterns and technical specifications and translating them into precise 3D recreations, accurate to the last design detail. This technology, Wilcox says, can reduce the time it takes to generate a true digital twin to potentially minutes, compared to the hours or days currently required for the creation of a virtual sample that is, from a product definition point of view, far less functional.

And second, once that digital twin exists, Wilcox believes firmly that its indelible links to all the ensuing processes of design, development, sourcing, and production make it considerably more valuable as a window into the additional metrics that designers are now being asked to consider:

“Having a digital product definition, and a genuinely interconnected dataset, makes it so much easier to see the impact of all your choices. If a designer makes a change, they can instantly see the effect it will have on costs, wastage, margin, sustainability and other variables, because every part of that relationship is drawing from the same set of product definitions. And just as importantly, if you’re defining products in a parametric 3D CAD solution, of which ours is the first to market, you’re making it far simpler for designers to actually make those changes to begin with.”

“Today, adjusting a design means having patternmaking skills, or working with a colleague or a supplier who has them, since this is the only way to really know how one alteration will affect other elements of the garment. Whereas working in parametric CAD allows an end user without patternmaking experience to change something for a subjective reason – adding a centimetre here or there to refine fit or to cater to a new trend – without requiring an understanding of what else needs to be changed to compensate. Iterating on a garment digital twin is exactly that: having the ability to make a minor alteration or even to swap out a collar or other component by simply modifying the garment itself, rather than working with a collection of pattern pieces that each have complex dependencies on one another.”

Parametric CAD is, yet again, an approach to 3D modelling that has a long and proven history in other industries. In essence, it represents a different approach to the creation of 3D assets – allowing end users to work with predefined features and sets of dimensions to quickly and easily control common variables such as size and position, and to swap out one pre-approved component for another. This is in contrast to direct modelling, which offers users more finite control through direct manipulation of geometry but which, as we’ve established, requires first-hand, brand-specific knowledge of how altering that geometry alters the underlying pattern.

Beyond the theory, a shift towards parametric 3D CAD in apparel and footwear (likely as a complement to, rather than a replacement for direct modelling) could have a host of benefits for creative teams who need to iterate and ideate at the same time as setting down clear product definitions that their colleagues and partners will then work with. By effectively setting guardrails within which those 3D designers can experiment, with reusable components and predefined parameters, fashion brands would be creating an environment that blends the best elements of creative empowerment without compromising on technical accuracy. And most of all, the organisation and the designer would share total confidence that any changes made to the garment will be reflected everywhere they need to be – without manual intervention.

This shift would represent a clear step towards bridging the communication breakdown between brands and their manufacturing partners, removing ambiguity and interpretation from the production process, and delivering full accountability by replacing flat technical specifications with clear, scientific instructions packaged within a digital twin of the product.

But the same shift would be truly transformative – and arguably downright essential – for localised, on-demand, digital manufacturing. As this approach begins to gain ground, fashion will find that the wiggle room permitted by the current model of production (where manufacturing partners translate technical specifications into actionable instructions) vanishes, placing even greater emphasis on the need for exact product definitions.

“Fashion has a clear mandate to move more of its production to on-demand because the industry currently struggles to calibrate supply to demand, which is why 30% or more of the garments we produce end up in a landfill,” Wilcox says. “From that point of view, there’s a very strong connection between waste, sustainability, and accurate product definitions, because garments that don’t fit as expected, or garments that were misaligned with design intent by the time they reached store shelves are the ones that are most likely to be returned, to go unworn, or to never be sold in the first place.”

“Actually delivering on-demand, onshore production is, as it stands, a much harder problem for the industry to solve, because its product definitions will need to be machine-readable if they are going to drive automated factories in small batches, rather than going through human-translated production overseas,” he adds. “But that sort of automation is absolutely where the industry is headed: rather than producing ten thousand identical garments, the new challenge is to produce that many garments that are all different in some respect, whether it’s personalisation through embellishment or embroidery, or custom fit that transcends set grading rules. That’s a huge topic for the industry to tackle, but it’s something I see as essential. And it’s something that the current approach to product definition simply isn’t equipped to handle, whereas it’s a core capability of a complete digital twin of a garment.”

Finally, thinking beyond production, the approach that Wilcox and his team at Clothing Tech are proposing could imbue consumer-facing experiences such as virtual try-on and interactions with 3D rendered garments with greater confidence. The consumer themselves may not know the difference between a 3D asset that was created for visualisation purposes and one that was generated to serve as a fully-defined digital twin at the point of sale, but they will recognise if the product that arrives differs from the digital version they were shown in form, function, or fit.

That type of breakdown in communication is what Clothing Tech is proposing to eliminate, because everywhere it occurs it carries a negative connotation for the brand. If it happens in-house, it becomes an operational problem at a time when few fashion organisations can afford inefficiency. If it takes places between the brand and their supply chain partners, it represents a risk to business continuity. And if the breakdown happens between the brand and their consumer, it becomes a crisis of reputation and loyalty.

In all of these cases, though, this is a problem that could potentially be overcome if fashion is willing to take a fresh look at digital product creation, and to rethink when, where, and how it’s defined.

This article was published in The Interline April 26, 2022. Clothing Tech is a member of SPESA.

SPESA members are encouraged to email news and releases to marie@spesa.org or maggie@spesa.org to be featured under Member Spotlights.