Milan
Research into fashion product development
Milan explores whether software could develop product concepts from a design team's references and conversations, ready for the designer to review.
Imagine a designer photographing a turtle shell because of its unusual colour. Later, during a collection meeting, they mention using that colour for an AW28 jacket. The photograph and the conversation belong to the same idea, but the connection may exist only in the designer's head.
The software would need to recognise that connection and develop jacket concepts from it. A useful interpretation would carry the colour into the jacket and the context of the collection, while leaving the shell's shape and pattern behind.
The material around a collection includes more than images: saved references, photographs of samples, meeting audio or notes, and conversations that give an observation its meaning. We're investigating how a system could draw on material the team chooses to make available, connect it and understand the direction expressed around it.
Arriving with something to review
The intended experience would begin with candidate concepts already prepared from those references and conversations. The designer's first deliberate interaction with the application would be to assess an idea that has already moved forward.
They could reject an interpretation, select a promising direction, change a proportion, ask for further development or decide that the original reference wasn't useful after all. The system would handle the interpretation and preparation behind the scenes, leaving the designer to judge the result and shape what happens next.
Understanding why a concept exists
An ambiguous comment could send a concept in the wrong direction. We're investigating how much context the system would need and how a designer could challenge its interpretation.
The person reviewing a concept should be able to see the reference that supplied the colour, the conversation that connected it to the garment and the changes made during review. That context would help them judge whether the proposal follows the intention or makes an unjustified connection.
We're also exploring whether comparison with supplied source material could flag a concept that remains too close to a reference. Such a flag could prompt the designer to intervene and develop it further. The comparison would cover the supplied material; it couldn't establish originality across every existing design.
Choosing what deserves to become a product
Once there are more concepts than a collection needs, the team has to select, reject, change and develop them. Which suit the brand? Which merely repeat the reference? Which have enough substance to pursue, and what's missing from the proposed range?
Those judgements would determine the direction. We're exploring Milan with the intention of giving designers more worthwhile ideas to assess and develop, with less effort repeatedly assembling references and explaining the same context to software.
Discuss Milan with Praxis