In the third and final instalment in a small series for WhichPLM, guest contributors, Nick Wei and Amy Yeung FY continue their exploration of 3D – with a look into the future. Nick has over 25 years experience in CAD/CAM and PLM; Amy has a strong product management background, with a Bachelor of Science in Textile and Fashion.
The early days of my [Nick’s] Apparel CAD career, back in the ‘90s, widened my vision of how the industry worked. It was also surprising to me how disconnected the supply chain was. Apparel manufacturing was not rocket science but the manufacturing process was very long and complicated. I got to learn about different types of garments, made out of different materials, and how these materials were made in different types of factories. Thus, the production process needed a very long lead time. It was fine at the time, as the business model then was much different than now – brands and retailers only had limited seasons in a year, and within the supply chain of the apparel world, they had to rely on a lot of middlemen, agents and brokers. The production slot was in a ‘made to stock’ mode, meanwhile, it was a sellers’ market where the cost of FOB was high.
The “good old days”: work with no time constraint
I worked for a fast growing company based out of Hong Kong, that developed a 2D CAD system for design. When it was acquired by a larger business in the late ‘90s, my role was automatically merged, opening up the entire spectrum from 2D design CAD to pattern making and all the way to 3D, CAM and PDM. I was amazed. PDM was what I envisioned the industry needed, as brands could use a solution that had seamless integration in development from ideation to production.
Now, almost two decades later, I believe there is still no seamlessly integrated system which can link up the entire process. There are many great systems in the market, such as 3D CAD, 2D textile CAD , 2D pattern making, PLM, and virtual fitting. But, the work process is not ‘fluid’ – files from one CAD system have to be imported to another and re-adjusted in order to achieve 3D simulation.
These technologies are designed to simulate a real garment, so that lead time can be reduced in the sample development process, but the simulation in each process is still siloed. Textile CAD is only used in textile mills and the simulation file has to be sent to the apparel designer to simulate the material in 3D, and, the textile CAD file has to be modified and readjusted to make the 3D simulation workable. Due to this, the lead time problem of the fabric/trim vendor is not solved. This vendor wants their textile or trim digital asset to be used by the designer so that they are able to understand what the market needs, thus, they can forecast what products to produce and reduce lead time and unsalable stock.
In my current job, I talk to apparel companies every day, who use technology but are still not able to achieve their business requirements. I want to share some examples.
A retail brand in Australia: “I am able to develop a range in 3D and allow my buyers to select styles and place orders, but it always depend on the available fabric we developed for them. It is difficult for me to find more fabric alternatives because there is simply not a pool of digital design assets available for me to select from.”
A shirt manufacturer in Hong Kong: “I have a 3D CAD system and I can show my shirt design to my customer. But when it comes to negotiation, the colour, and plaid size changes which leads to cost changes due to marker wastage and colour minimum. I always had to spend time communicating with the fabric mills to confirm the price to offer my customer; and I may lose my margin or my order when I am not able to react quickly. I need a system that can integrate textile design, marker wastage count with 3D CAD which can reflect costing!”
A fabric mill in China: “I have a lot of overstock that sits in the warehouse unable to sell. How can I sell my overstock fabric? In fact, the reason I have overstock fabric is because I was not able to forecast the market demand. I wish I had digital textiles the designer could select from, so that their buyer would be able to use my digital fabric to simulate their style and sell to their buyer, when they buyer can make a faster decision. Then, I would know what to produce earlier and be able to forecast my production accurately.”
The irony is: apparel technology in 3D and 2D simulation has been available for the past 15 years, but these technologies are not able to fully integrate. Yet digital supply chain is still disconnected.
Let’s take a look at the product offering in the 3D CAD eco-system:
Diagram: ideal scenario of digital asset shared for virtual sampling
There are some 3D CAD suppliers who develop their systems from 2D pattern making bases, where they have their own 2D pattern CAD as the foundation. And, there are others who develop from animation. Both have their advantages and disadvantages. Thus, different vendors have different philosophies in their development directions based on their fundamental technical competency. Some companies own both systems: one for making nice simulations of their new designs to show to their buyers, and one for making 3D which can produce real, manufacturable production patterns.
Let’s discuss some other products that complement 3D CAD:
- 3D avatar development is a critical part of simulation since it orchestrates the pattern and fit of the garment. So far, I believe there is only one vendor in the world dominating the industry. the digital avatar is already compatible for all 3D CAD vendors.
- Material physical property measurement for tension and shearing thickness will affect the drape and fit of the garment on the mannequin. Different 3D vendors have different equipment for that, but should that physical property information come from the fabric supplier? And should the fabric’s physical property information be a standard across the industry? Currently, different 3D vendors have their unique tools to test that measurement.
- Fabric and Material scanning. After a 3D sample is completed in simulation, the seller may want to zoom into the item to see the fabric quality or construction. If the fabric is not simulated by the textile CAD system, the material visual quality will not be high enough for visual inspection. They may need to do high quality, real fabric scanning so that the buyer can review the detail. There are different technologies available for this, and some use high quality cameras for scanning whilst others just use a mobile phone, but all have built their own ‘light box’ for scanning. The quality of the scanned material varies due to the scanning device and the software algorithm. This fabrice service texture image should come from a fabric mill, right from the source.
- Body scanning is an important device to create the avatar and there is an increasing demand for the made to measure industry. There are a lot of vendors providing such solutions; some focus on apparel, and others on healthcare. The technology has developed rapidly, from body heat detection to laser, then later to camera and now using Kinect. Some newer systems are done via mobile apps. If it is to be used for the made to measure industry, how will ir work with adjusting patterns? So far, a real auto pattern making system based on scanned bodies is not yet a feasible technology.
- Virtual try-on technology was a big initiative for the online shopper. The objective of such technology is to put your body measurements or your own scanned mannequin into a virtual shopping environment, then you are able to try on clothing and check how it fits. The business rationale behind this technology is that, if you are able to find the right clothing at the right size, then the chance of returning or refunding will be less. It’s a technology ideal for complementing 3D CAD and body scanning so that the ‘scanned body’ can try on the true fit virtually.
- Animation and rendering technology enhances the virtual simulation experience, allowing B2B or the end consumer to view the product in a more ‘realistic’ environment. For example, if the vendor is selling a golf wear collection, the buyer can see the item modelled in a golfing pose, on a golf course. The buyer can see the simulated fit. Some 3D CAD vendors already embed such features but others don’t. Should these animations and rendering environments be in a ‘predefined’ platform so all CAD vendors can use them? No one should re-invent the wheel…
- The objective of 3D Store Displays is is to allow the range planner and store manager, as well as the store display designer, to standardise the display of the store, such as defining which product should be put on which rack, and how the display should look in a 3D environment. However, 3D store display systems are usually siloed from the digital world. What if it could connect with a 3D virtual store, allowing the buyer to have an in-store experience?
- Virtual Reality is a system/device in which you wear a pair of goggles to see a virtual environment. Anyone who has watched the movie Ready Player One knows what it is; it is widely used in the gaming industry and is apparently a somewhat controversial topic in the apparel industry. Some VR vendors have already developed virtual shopping environments to enhance the online shopping experience. Should the 3D store experience be connected to VR to enhance the virtual shopping experience? The end consumer could pick up their favourite garment in the virtual world and try it on. This technology is currently in the conceptual stage, and not yet fully commercialised.
- Product Lifecycle Management (PLM) is, in fact, the platform that should tighten the output of all of the above technology, linking up the entire development process from merchandise planning to story board creation, the initial design sketch to tech pack, costing, sample stage pattern file, with sizing, avatar information, 3D files and all the virtual and physical iteration of the style. Material information of fabric, trim hardware should also be store in the PLM platform for the designer, pattern maker and 3D designers to use. Some PLM vendors are already able to import 3D in the system for revision and design collaboration purposes, but once the 3D has been modified, the 2D pattern files have be updated for actual production, and costing should also change. There are still a lot of areas of disconnect to make the entire process more ‘fluid’.
- 2D pattern CAD is a mature technology that almost every apparel company use There are plenty of vendors in the field, and the good thing about this field is that the file format already has a universal format – AAMA DXF – and it allows the 3D technology to use.
- 2D textile CAD is also a mature technology for textile design, and there are variety of vendors focusing on different types of textile. The basic ones are print design, basic weave design, and simple knit design but there are vendors that specialise in crochet design, lace fabric design, sweater knit, carpet, jacquard, and tricot. These digital assets are not shared for the 3D designer to use.
- 3D hardware CAD is an important tool that renders the hardware in apparel – such as buckles, zippers, metal or plastic badges and other hard items. The technology for hardware rendering is mature and it has been used in multiple industries already. However, the hardware vendors in the apparel industry do not yet commonly share their 3D files with apparel designers.
- Artificial Intelligence in the apparel development field is still an early stage, but there are several areas it should address within the 3D eco-system. One is auto stitching – which means, when a 2D pattern files are received, A.I. will have the intelligence to recognise where to stitch the seam line and that will reduce the time of 3D simulation. There are start ups that have developed algorithms to forecast fabric sales, and style sales based on big data in population change, weather, retail sales index, and POS statistic. Thus, it will help the vendors to forecast and pre-book materials and vendors at an earlier stage.
Diagram: Digitisation of the supply chain
The good news is: we have a full set of technology to build the apparel digital world. But, will the industry really able to transform? The issue is how all of these vendors will work with each other, and how these technologies will integrate or at least build a standard format so that all information can be communicated, decisions can be made faster, costs can be reduced and wastage can be reduced.