Yesterday's post closed with a wish – seeing more traditions from the past being rediscovered and reinvented for the future also with the help of technology. Yet, in many ways, this is already happening and at times it doesn't only involve the design process, but also the research phase. At the beginning of March, for example, Microsoft announced it had devised an Artificial Intelligence-powered system for Emily Adams Bode, founder of the menswear line Bode.
The New York-based designer has become known for reusing antique textiles or techniques like quilting (employed to replicate vintage patterns) in her modern menswear garments to create unique pieces. Bode sees the possibility of repurposing fabrics in new designs as a way to add a narrative value into each of her pieces.
To make sure her consumers got precise information about the textiles employed to make a specific garment, Bode and her team would add a handwritten hangtag with notes about the fabric and the design, but their process has now become easier thanks to a system designed by Microsoft.
Built with Microsoft Custom Vision AI and an integrated database system called Cosmos DB (extension of Microsoft Azure), Bode Vault is a digital quilt expert that was developed over the course of one year.
The system allows Bode's sourcing team members to take a picture of a textile with a tablet or phone, upload it onto the app and immediately get a profile with its name, country of origin, style, approximate time when it was made, artist or maker, colours and materials.
The app also provides similar patterns and added information like intriguing stories about that specific motif. The entries can be edited, obviously, in case further information must be included.
The AI-powered digital quilt archive has a historical and a functional purpose: it is an archive of historic quilts and can therefore be conceived as an educational tool, but it is mainly used to keep track of inventory and identify stocks available (a key point when you want to make a garment with antique textiles and bespoke materials as in Bode's case and you must rely on the quantity of material you have or you can find).
The AI was trained with a library of images put together from Bode's own quilt collection and archive and with information provided by a team of professionals gravitating around Bode, among them textile experts, historians and American quilt dealers around the country. Quilt pattern recognition is a time-consuming job and this is one of the many processes the AI can certainly help speeding up.
For the time being this tool is only available for Bode, yet this fabric library that matches human knowledge with machine insights opens up other possibilities for different brands, but also for textile archives and museums.
Last but not least, the application would be really useful during situations like the Coronavirus lockdown: such online systems would indeed allow students, researchers and designers to keep on working remotely on their personal projects without suffering from major disruptions.
Rispondi