Fashion goes three-dimensional, and the law wants the facts

For most of the last decade, 3D in fashion has been judged by how a garment looks. Can a buyer approve a colourway from a render? Can a designer try a sleeve without cutting one? Those are real gains. But between 2024 and 2027 the question that matters is changing. European law is turning each garment into a record, with facts attached about its fibres, its repair and its end of life. A 3D garment that is only a convincing picture serves the first job. A 3D garment that is a measured representation of the thing can serve both.

The rules arrive in sequence

The European framework for textiles is now a timetable rather than a proposal.

The Ecodesign for Sustainable Products Regulation sets the frame. In April 2025 the Commission adopted its first working plan under it, covering 2025 to 2030, and named textiles, with a focus on apparel, among the priority product groups, alongside steel and aluminium, furniture, tyres and mattresses. The Commission's textile page puts the planned adoption of the textiles delegated act in Q4 2027 and notes that the timeline may move as the work progresses. Its FAQ adds that operators will then have a transition period of at least 18 months before requirements apply.

What a textile Digital Product Passport will hold is set by that act, but the Commission has listed the likely categories: product identification and characteristics, fibre composition, information supporting use, repair and maintenance, information for reuse, resale, disassembly and recycling, origin, the economic operators involved, and documentation for compliance and traceability.

Two other rules already apply. Since 19 July 2026, large companies across the EU may not destroy unsold clothes, clothing accessories and footwear, with narrow exceptions for products that are unsafe, damaged, counterfeit or refused by donation schemes; medium-sized companies follow from 2030. And the revised Waste Framework Directive, Directive (EU) 2025/1892, entered into force on 16 October 2025. It creates extended producer responsibility for textiles across the Union: producers of clothing, footwear and household textiles pay for collection, sorting and recycling, and Member States must have schemes running within 30 months, with fees that may be modulated to discourage the most wasteful practices.

The pressure behind all of this is volume. The European Environment Agency put EU textile consumption at 19 kg per person in 2022, up from 17 kg in 2019: 8 kg of clothing, 7 kg of household textiles and 4 kg of footwear.

What the rules reward

Read as engineering requirements, these rules reward three things.

The first is fewer things made that nobody wants. A ban on destroying unsold stock moves the cost of an unsold garment onto the company that made it. The cheapest unsold garment is the one never produced, which pulls decisions about fit, fabric and quantity earlier, before cloth is cut.

The second is facts that travel with the product. Fibre composition, repair information and end-of-life information have to be true of the garment as made, held in a form a system can read, and kept current as the product changes.

The third is accounting for the end of life. When producers pay for sorting and recycling, what a garment is made of, and how its parts come apart, becomes a cost line.

Each of these is easier when the garment exists as an exact digital record before it exists as cloth.

Where 3D design stands

Digital sampling already does part of this work. Fitting a garment on screen saves physical samples and the weeks it takes to make and ship them. The harder part is reaching everyone else. A 2025 study of small London designers presented at the IFFTI conference found clear gains in iteration speed and sampling cost, along with steep barriers: around £2,000 in computing hardware as a minimum, and six to twelve months for a designer to become proficient in the software.

There is also a gap in what a typical 3D garment is. Most digital samples are built to be looked at. The fabric is an appearance with drape settings tuned until it looks right; the body is a standard avatar; the pattern may or may not match the one that goes to the factory. That is enough to approve a colourway. It is less useful as evidence. A fit judged on screen is only as good as the body, the pattern and the cloth behind it, and a passport entry is only as good as the record it was drawn from.

Exact representation as the shared record

The way through is to treat the 3D garment as a representation with measurements attached, not as artwork. The same record then carries the pattern pieces that will be cut, the body they were drafted for, the fabric with its measured properties and its true repeat, and the fit read off the draped result as numbers. Composition and construction live in that record because the simulation needs them, and the passport can take them from there.

Two things follow. Design decisions move earlier, because a fit that can be read as ease at bust, waist and hip in centimetres can be argued about before a sample is sewn. And the claims a brand makes downstream, about composition, repair or recyclability, come from the same record the designers used, rather than a spreadsheet maintained beside it.

The weak point is the cloth. A drape is only as true as the fabric properties behind it, and those have to be measured on a bench, not chosen by eye. Until they are, every simulated fit carries an honest caveat.

Where Livgens fits

Livgens represents the thing exactly, in sound, light, matter and motion, and Living Worlds runs it forward through its lifecycle. Livgens Render works with surfaces, objects and space, and drives Blender, Unreal Engine, FreeCAD, KiCad and RealityScan from one conversation. In a published case study with a fashion house, Render built a body fitted to four measurements (height, bust, waist and hip) to within a tenth of a centimetre, drafted a parametric pattern for it, and draped it on the GPU with NVIDIA Warp until the cloth came to rest on a 1 cm mesh. The fit was read off the draped garment as ease at bust, waist and hip and stretch along every seam. A shirt draped in 34 seconds instead of 208 on the CPU, with the fit reading the same: 6.1 cm of ease at the bust on the GPU against 6.5 cm on the CPU. Prints and weaves are laid at their real repeat, and the browser view, a ray-traced still and a live Unreal Engine stream all draw the same settled garment.

The case study states its limit plainly. No fabric has yet been measured on a bench, so every drape is shown as unvalidated until it is scored against real cloth; fabrics measured on the bench are listed as next. Render is in early access.