As part of our ongoing ‘Ask the Experts’ series, we have brought together some of the Fashion industry’s leading Digital Transformation specialists to answer your most pressing questions. Today, we ask the team: what are the challenges and opportunities in applying automation to digital prototyping?
Here’s what they had to say…
When asked to streamline a process, people often confuse the symptom with the cause. The need for automation is often due to a flawed process or a flawed user experience inherent in the technology. When people ask to “automate digital prototyping”, it’s likely a combination of both.
Let’s first ask the question, “why do we want to create a digital prototype?”
Digital prototypes have many potential uses:
- Create a better physical prototype, much faster (fit approve on first physical)
- Quickly validate understanding of design intent
- Provide a fast visualisation for merchandising decision-making and pre-selling
- Reduce physical footprint, waste, and carbon emissions inherent in shipping
- Get closer to market and reduce prototype/sample costs (if paying for samples)
- Create content for consumer insights
Based on these business goals, how would we want to improve our business process and technology experience more easily achieve those goals?
One way is to provide a better, more connected infrastructure that allows the team to seamlessly share information between merchandising, design, product development, the factories, and sales. Solutions like VibeIQ, Dtail by Pixelpool and Centric VIP offer this capability. Having a collaborative environment for all roles to view and make decisions on content is likely more critical in our workflow than simple tool automation.
That said, the ability to quickly create digital prototypes at either the home office or the factory will aid in faster decision-making and team efficiency. AI-based solutions like Synthesis from Six Atomic and Fashionr from Clothingtech allow teams to quickly go from concept to manufacturable data more quickly while maintaining block-based fit standards.
While these solutions can provide speed, be sure to ask how, or if, your business is prepared to make use of that speed. If there’s not a willingness to reduce the calendar or use this methodology for quick-turn product, for example, the efficiency benefit will end up being for a very narrow segment of your organisation. If that’s the case, the ROI will be hard to justify, which is why speed and automation should be in context of how it will benefit the entire organisation and process.
Reducing the friction for creating a digital prototype is always going to be nice to have. Creating an organisational capability to make use of the digital prototype for better decision making and reducing time to market are must-haves, and will be the key to success for any digital transformation effort.
Automation of routine work is always welcome; however, there is very little that we do in the creative space that is “routine.” Fashion means new and from my Differently Enabled perspective, when you add the complexities of asymmetrical/atypical body morphology, and exacting Raw Material requirements based on the customer’s lifestyle, very little is “routine.”
One of the big unlocks to automating our manual madness is interoperability. PI Apparel and the 3DRC are addressing this, and they will tell you “the struggle is real.” I’ve been banging on for decades that our technology collaborators must stop adding patches to the patchwork quilt we call a technology ecosystem, and instead look at how they can contribute to the flow of data and images that result in physical product. And the data and imagery needs to be from past related Finished Goods and Raw Materials, as well as the current ones we are developing. “Stop gap technology” needs to stop and the “one stop vendor shop” is not the answer either.
As AI matures in our sector, perhaps we can start leveraging that to find the larger workflow repetitive patterns we are currently overlooking? But only when ALL of our work is digitised can we truly leverage this type of AI. We are a few years away from this now.
I am working on bleeding edge tech with a number of young start ups, and I am optimistic that we are moving in the right direction, but only when we unite as an industry to transform ourselves, will great things happen.
So if you’re reading this, give yourself a pat on the back for being part of the greater good! Attend your tech partners Customer Advisory Board meetings, continue to have a voice at conferences, share your success and failures.
Congratulations for being part of the change!
Find Craig here.
Digital prototyping in most programmes is still a surprisingly manual process. I think it’s clear to everyone that manual work = time, and automation = less time. It’s interesting to observe key players approach the topic from different angles.
As an example, let’s look at the programmes CLO3D (by CLO VIRTUAL FASHION) and Fashionr (by clothingtech):
- CLO3D has been around since 2010, while Fashionr is comparatively new
- CLO3D has a large variety of tools and functions, and can be used to achieve anything from simple prototypes to completely photorealistic garments. Fashionr, on the other hand, has the goal of automating the prototyping phase.
- With CLO3D, you can cover a large scope of options, roles, and phases, while with Fashionr the goal is to simply load in the technical documents (DXF pattern, tech pack, choose garment type etc.) and let the programme assemble the garment for you. It also has further automated steps, like choosing a collar type from a list and letting the programme execute a series of steps, but isn’t meant to be used for high-end visualisation.
As a 3D Artist, I appreciate the flexibility of CLO3D, and being able to take one asset further into multiple parts of the value chain, but the automation of Fashionr has its own appeal.
As cloth simulation evolves, it’s worth watching in which directions the different programmes move.
Find Sophie here.
I believe our industry has an opportunity to leverage new methods of make, including: additive manufacturing; engineered knitting; 3D printing; non-woven material applications; new bonding technologies; and more, which will undoubtedly take full advantage of the digital nature of the 3D artefact (whether it is a shoe, a garment, a bag or anything whose manufacturing and visualisation can be driven by a digital asset).
Granted, there is still a lot of work to be done to close the digital loop and to enable product to be prototyped and even manufactured in a completely automated way. Nike (through their partnership with Flex) and Adidas (with the Speed Factory initiative), among others, have paid the price for being a bit too early to the game but we’re getting excitingly close to the point where automation is a financially viable option.
Personally, the biggest unlock I see is the ability to create batches of one. The Holy Grail is to be able to design a garment or a shoe, create powerful algorithms that would decline this design based on a number of anatomical and behavioural factors (such as an athlete’s foot shape, gait, range of motion, need for zonal protection or breathability), create a product that is unique to those individual factors and feed a manufacturing process that spits out a finished product within a short period of time (ideally within 30mn).
Yes, this is more than prototyping, but I believe that with the right technologies, the boundaries between prototyping and manufacturing will be blurred, replaced by an opportunity for constant improvement and adaptability.
Thus, contrary to popular belief, automation could and should lead to bespoke, highly customisable products of higher value, instead of cheaper, identical items that are the result of the pursuit of the lowest cost possible. Personally, can’t wait!
Find Safir here.
Automation in digital prototyping would be an industry-shifting revolution. Unfortunately, automation in prototyping today is like automation in cooking. The best we can do is build a process to create a rigidly consistent product. Hello processed foods, goodbye home-cooked meals, restaurants, and haute cuisine!
We live in a time with so much automation that it is easy to forget that automation is just programming. When you ask your phone to “call mom”, it doesn’t trace your genealogy or use a process of elimination to define which is the one you call ‘mom‘. In 1939, “automation” was asking the operator to connect you with “New York, Eldorado, 53598.” Over time, we programmed the switchboard operator away and now Siri needs only to find the number you’ve programmed as “mom”, Still, I do sometimes still get connected to MOMA (The Museum of Modern Art)…
Initiatives to automate digital prototyping have made strides but are still quite limited. Let’s remember computer programming as we know it today is MUCH younger than garment making. Jacquard clocked an early win for computer programming of textiles in the 19th century, but we were covering our bodies long before that. Therefore, garment making does not comply with rules and standards convenient for programming (and automation).
Still not understanding the complexity? Okay, let’s look at sleeves. How would you instruct a computer program to recognize the front and back of a sleeve? Some countries put a double notch on the back, some on the front. What about all the different sleeve variations (set-in, raglan, 2-piece, dolman, cap, puff, bell, etc)? Remember the silhouettes also continue to change over time. We can add ease, pleats, gathers, and seams as we choose, and the pattern shape can also change with fabrication.
The innumerable variations pose challenges for automation and machine learning.
The apparel industry has seen successes with made-to-measure and pattern wizard automation as they’ve been around for a long time. Unfortunately, those tools require brands to have a deep understanding of how their product is engineered from scratch and modified into new designs. And sadly, much of that knowledge has dissolved since these tools were invented.
Moving forward, brands focused on automation should re-establish that garment engineering prowess and add some software engineering to the mix. Then, if our digital prototyping tool providers continue to support configurability and APIs, I can see a future where tools that automate digital and physical prototyping deliver ground-breaking results.
Until then, automation can only get us part-way there!
Find Christian here.
