3D body scanning question

Excerpted from my mail:

Another topic I’d love to hear your take on: body scanning. Specifically, the goal of taking a 3-D scan of an individual and having an automated system whereby that scan drives the appropriate adjustments to a base pattern and turns out a customized pattern for that individual on the other end. Is it do-able in a commercial sense? Perhaps not for a company like H & M who turns over styles super fast, but could a company such as Lands End take some of their most popular designs and make them available for such a customization service?

I’m probably not the best person to answer this question for several reasons these being:
1. My focus is basic education using basic technologies; I’m a lean manufacturing proponent. For example, I don’t write about CAD either and 3D body scanning is a leg up from that. The reason is that technology -while useful- is not a substitute for core competencies. In the case of CAD, you need to know how to make patterns by hand before you can effectively use the software to make patterns. In my opinion, most of the problems in our industry are due to a lack of competency. Somebody’s got to cover the basics; hence my mission. I do bring in outside authors for given topics (one is upcoming) but I keep a core focus on the basics rather than the latest and greatest.

  1. I am conservative when it comes to adopting technology. The potentiality of technology is too often used as a magic wand or a miracle cure. Toyota, for all its exemplary products, is similarly conservative. A major concept of lean manufacturing is that core competencies must be developed using standard tools first. Once a technology is proven, then it is cautiously adopted. This hasn’t hurt Toyota in the slightest. Rather, it’s prevented them from becoming yet another dead automaker. Another example is JC Penneys with regards to the big push to adopt RFID chips. Penneys is “watching it closely” but remains unconvinced. Similarly, there’s always the latest and greatest in manufacturing philosophies laid similarly by the wayside.


3. As time passes, I’m becoming even more convinced that niche manufacturing will continue to stratify and grow. While I don’t deny that retailers are increasingly converging, the increased demands from consumers will continue to drive the impetus of niche manufacturing. For example, the latest news from the NY Times is that Neiman’s, Saks and Bloomingdales have dropped petite sizes from their stores. Considering that the average woman is a petite being 5’3″ (according to the latest data); I can only see the logical consequence of this as driving yet more demand from consumers to create apparel to fit this market. When you couple the long lead times required for push manufacturing schedules, this will only serve to heat the demand for faster turn around -in other words- lean manufacturing from smaller companies. I have little doubt big box retailers will continue to fight over a smaller and smaller portion of the American apparel pie, leaving yet more opportunity for DEs to develop profitable enterprises whilst retailers scrabble and scuffle. I have long predicted that the total number of apparel companies will increase, reflecting the increasingly diverse needs of apparel consumers. Returning to your question, I don’t see the lack of 3D body scanning as being an impediment to the growth of DE companies, any more than their lack of CAD is. The latter need can be met by pattern services in the grading and marking process.

  1. With specific regard to your question:

“Perhaps not for a company like H & M who turns over styles super fast, but could a company such as Lands End take some of their most popular designs and make them available for such a customization service?”

I agree that fast fashion companies -what I’m trying to build here with the focus on lean- will remain less affected. For one thing, they’re not focusing of fitting broad swathes of the market but niches as defined by income, interests and increasingly, age. Second, their cycle time is faster, consequently their fitting cycle iterations are reduced to a season or at most two. So, I don’t believe they’ll need to rely on having such technology in house. With regards to Lands End, who can say? As a large enterprise and a push manufacturer, their constraints are their very structure so I’d haphazard a guess by saying it is unlikely. To adopt 3D technology, their entire infrastructure would need to change. First they’d need accessible entry points to measure consumers meaning a lot of 3D stations. As a catalog company, I’d find their options for entry points to be limited at best. Second, they’d need rapid cycling in patterns (read: a lot more pattern makers) to say nothing of CAM (computer aided manufacturing, costing lots and lots of money), meaning they’d need the capability to cut one offs non-stop and at the drop of a hat. Third, they’d need domestic production, meaning they’d need a lot of small operators to sew this stuff up. I just don’t see it happening. Their very size constricts them. Lands End may realize profits on their existing economy of scale but I can’t see them transitioning downward successfully. If I’m not mistaken, Levi’s tried this -an early adopter of new technology- but has since abandoned the enterprise.

Personally, while I find body scanning to be interesting, I don’t find it viable for most DEs. Not to say that couldn’t change in the future but I don’t forsee widespread adoption of the technology. I think it could be useful for manufacturers in the development of prototype fitting and in the development of targeting a given consumer demography but little beyond that. That’s not to say some DEs won’t adopt it, indeed several have. Just that adoption will be limited to those with deep pockets or those who have backers with deep pockets. Having the technology is only part of the solution. One must also have the people who can translate those scans into patterns and also, one must have the one-off cutting capacity to say nothing of the equivalent of a factory full of sample makers sewing up the one-offs. From outside the industry, this issue has been seen as a possibility to go truly lean but I don’t think it’s appropriate. As I’ve always said, you can’t pull corn or coats like cars. In dirt industries -of which apparel is one- batch processing in some degree will remain unavoidable.

Get New Posts by Email

10 comments

  1. With regard to the sort of technology you suggest, if the question is ‘is it doable’ or ‘is it possible’, of course the answer is yes. There is nothing in the process of body scanning, downloading to a fully automated cutting and sewing operation and cranking out something precisely to your size very, very quickly that is beyond the current realm of manufacturing technology. That is the standard answer to just about any manufacturing technology question today.

    The issue is cost. The investment required to automate that process would be enormous. It isn’t being done and is not likely to be done soon because a shirt produced by such technology would cost tens of thousands of dollars.

    Just about anything that is manufactured can be automated. The question is always whether it is economically feasible to do so.

    As Kathleen pointed out, the niche manufacturers don’t do the volume necessary to justify an investment in such technology. The high volume people are all in a mass production/ one size fits all (or few sizes fit most) mode and are not interested in customization.

  2. Alissa says:

    Levis tried this a while back. I remember seeing this in a San Diego Levis store. Here’s a link to a Feb. 2001 article in Fast Forward
    (see the 8th-10th paragraphs) describing an unsuccessful user experience with the technology in their San Francisco store. I don’t know if they’re still offering this, but I didn’t find it in a quick look at the Levi’s site…

  3. Kathleen says:

    oooh thanks Alissa! Perfect! From Alissa’s source I quote:

    An hour and a half later, after donning a specially-designed Levi’s bodysuit so that the lasers could read my contours more efficiently, and four or five attempts to map my body shape in an enclosed steel room straight out of The Matrix, reality had a cruel, cold laugh at my expense when the beleaguered salesman and I determined that Levi Strauss is not a friend of high-waisted women. He finally, sheepishly admitted that scores of other women had the same problem. This was partially the result of Levi’s redesigning their custom offerings to include men’s jeans and eliminating measuring the rise in addition to the waist and inseam.

    It would appear that even with body scanning, well fitting apparel is not guaranteed. It all goes back to implementation -which if you think about it, is no different from manual measuring.

  4. Esther says:

    All of the major CAD developers are pushing this 3-D technology in hopes of speeding up pattern development (specifically draping, and ultimate customization). I have yet to see the software live up to that goal (and I doubt it ever will). It is exciting technology and it will eventually become affordable, but is it really needed?

    There are still several problems the engineers need to overcome. The body is not a rigid 3-D structure. It is soft and squishy. Fabric is fluid. And each individual, fabric, and comfort level is going to vary too much. That is why there is such a thing as standard sizing. These are only some of the many variables that have yet to be adequately reproduced digitally, and still save time.

    On top of that it would require designers and patternmakers to change their whole thought process. I am sure the same arguments were made with the early beginnings of CAD, but in this instance I see even more difficulty. Designers like to design with their senses, including touch. To place a piece of fabric over a form or body requires a set of skills that do not translate easily to a virtual environment. While patternmakers are more technical, they still need to see things in fabric. Perhaps a Star Trek Holodek would solve the problem?

    A skilled patternmaker can work in a 2-D environment and see the 3-D effects in their mind. IMO, the 3-D technology would not shorten the development time. I will wait until I can truly test this in a real world environment to give my final opinion. In reality I think this is still has 20-30 years before it becomes a viable option.

  5. Claire says:

    The question about body scanning came from my email. My curiousity has less to do with whether or not the technology will be implemented due to costs, but is the technology capable of working in the first place. The commercial goal seems to be not to use the scanner to draft patterns, but how do the measurements of the body drive idiosyncratic and customized alterations to existing base patterns?

    Not that the scanner does not work to take a perfect scan of the body (except sometimes the crotch, but more on that in a minute) but as we all know the inside ‘shell’ of a constructed garment is not exactly the same as the outside shape of your body.

    Full disclosure – I was recently an Apparel Design Masters student at Cornell University where body scanning is a focal point of the department’s research. While it was not the focus of my own research, I did participate in a graduate level class where our semester long project explored the process of having a 3-D body scan drive automated, yet customized, alterations to a base pattern. Nike was our corporate sponsor and they provided us with a fully graded set of sizes for one of their most basic pants – an elastic waist woven track pant and sewed the samples for our ‘subjects’ based on our automated patterns. It was as simple and straightforward a pants pattern as possible; two front pieces, two back and the top turned over as the elastic casing. The results? Many subjects’ pants fit great, others not as good and we had one or two humorous disasters.

    My team was assigned the task of deciding which measurements would drive which kinds of alterations to the pattern. We poured over every pants fitting book we could find. We had 5 measurement changes to make, but each one was a one to one cause and effect. They could not interelate with each other. The low hanging fruit was the hem alteration, of course. The crotch curve, however, was a whole different matter. Even if the scanner did capture an accurate crotch curve (if not, a manual ‘crotch patch’ was required and besides defeating the purpose of automation, doesn’t that make you giggle?) it was hard to make a simple automated change to that curve based on one measurement. (we chose crotch depth as I recall – a measurement that the scanner can do but a human cannot – at least not with a tape measure).

    Another difficulty when trying to pull measurments from a scanned image is not the accuracy of those measurements, but having the computer program know where to establish landmarks and take those measurements. The waist is tricky because not everyone’s is clearly defined. (not to mention everyone has a different idea of where the waist should fall on a pair of pants). But the base of the neck seems to confound the measuring software regularly.

    My own patternmaking skills I would still consider to be novice. (my research went off in a cultural studies direction). But in my own costuming type of background my experience has been that fitting a base pattern (especially a complex one) to an odd shaped body (and lets face it – most bodies are) seems to involve a certain kind of ‘knowing’ and craftsmanship that comes with years of experience. If you took 5 top notch pattern makers and gave them the same fitting ‘problem’ would they all produce the same solution? How much of this is human artisanal work and how much of that process could reasonably be automated?

    I’d love to hear some opinions from professional patternmakers on this.

  6. Jane says:

    In response to your request for opinions from professional patternmakers (I am always willing to express mine)regarding the statement:

    “If you took 5 top notch pattern makers and gave them the same fitting ‘problem’ would they all produce the same solution? How much of this is human artisanal work and how much of that process could reasonably be automated?”

    There is alway more than one way to skin a cat. (no offense to cat lovers, I have 2 of my own) Certain types of fitting problems have basic alterations. However, many others may have more than one way to correct a fit problem. I believe all fit corrections can be done on a CAD system so that the base pattern has a good fit on the form or fit model.

    However, I have no experience in body scanning to know to what level of customized fit the system is attempting to achieve. Is the product attempting to achieve a total personal fit garment (like going to a dressmaker or tailor) Or is the goal a semi-custom fit (getting the right waist or hip or correct inseam length) There is a lot of distance between these two different scenarios.

    I began my career doing custom work and that experience has been so helpful in learning fitting and its solutions. I then went to work in the industry and used those solutions in problems that arose in production work working with hard patterns. Learning about the relationship of the various pieces to each other that is required to get a clean hanging garment. Finally,I learned to utilize both fitting and hard patternmaking skills in computer patternmaking. Each skill bringing experience from the one before.

    I totally agree with Kathleen’s remark that a patternmaker needs to know the craft in its basic form first. This includes knowing how to fit. Making 2 dimensional pieces, smoothly cover a 3 dimensional body. I have worked with patternmakers that have neverdone any fitting or made hard patterns in the industry and it shows up in an ill fitting finished product. Just because a pattern piece looks good on the screen, does mean it will look good when sewn to the other pieces and is hanging on the body.

    I really don’t understand the goal for the body scanning concept. Is it to create individual products one at a time or is it to gain a better understanding of average shapes and to translate that info into mass produced product?

  7. Stacy says:

    I am a professional specializing in the use of 3D CAD and 3D whole body scanning technologies. It is easy to mix the 2 concepts up, as they both use 3D visualizations so lets break it down.

    2D to 3D – Many Fast Fashion and innovative DE’s are using this technology for 3 reasons:
    1. Virtual Prototyping – With a proper 3D fit model, a pattern maker and spec designer can test out their designs virtually prior to sewing a sample. For many companies using offshore and domestic production, up to 3 to 4 samples are required for approval for production. As all these samples are shipped via air/ground, the expense and time of producing samples can be saved (material and labor). While real samples still need to produced, the first or second sample can be constructed virtually at a point when design and general fit decisions are being made. A Fit Specialist/Pattern Maker applies fabric and stitch properties to closely simulate the expected look of the garment. What does this really mean for skill base in North America? Pattern makers will still be needed at these companies as these skills are required for effective Virtual Prototyping. Basic foundation hands-on skills and experience is important before you can work to a 3D environment.

    2. Specification Packages – The Design Team now has a 3D view of their design that can be emailed or included in the manufacturing spec package. As we all know, a 2D image provides a lot of construction information – imagine what a 3D image can do. The apparel manufacturing is one of the last industries to adopt 3D technology for product development.

    3. Merchandising and Marketing – With the quality of images that can be achieved with 3D technology, these images can be used for catalogs, presentation materials and website presentations. Show your skirt in five different color ways/prints, reflecting the drape of the fabric and allow the consumer/buyer to rotate the image to get the best view. What are the advantages? The cost of samples and a photo shoot vs time to prepared 3D images using already prepared patterns.

    Fit Model – Most 3D CAD companies offering 2D to 3D solutions use a parametric model. This means that the model can be ‘morphed’ to your Fit Model measurements. In some cases, the model can not be morphed exactly to the Fit Model measurements due to posture or specific market requirements. The alternative is to have your live Fit Model be scanned. This electronic information can then be converted to a Fit Model that can be used in 2D to 3D CAD solutions.

    Made-To-Measure (MTM) – Made to Measure has evolved over the centuries! In North America it is not as popular as in Europe and Asian markets. CAD technology became used for MTM suits about 20 years ago. Yes, it has been in use a long time and by some very large and small suit manufacturers. Their approach has been for the tailor to measure the customer, recommend a starting size and fit the jacket. From the fitting the tailor will identify alterations required. These alterations are then inputted into the CAD system against a pre-programmed pattern along with fabric and construction order information.

    What will whole body scanners do for the process?
    Whole body scanners provide the raw data necessary for Extracting Measurements. The definition of the measurements extracted is up to the pattern maker/CAD technician as required to alter to the style. So now the retailers offering MTM can assist the tailor in taking accurate measurements and also have software to recommend the starting size. These extracted measurements can also be applied to a pre-programmed pattern to alter the shape of the pattern to fit the customer. In some companies, they do not have the customer try on a size, but just use the extracted measurements to alter the base pattern to fit them.

    What are some of the success factors?
    1. The base size pattern fits the “fit model” perfectly. the “fit models” measurements are used as a foundation for programming the patterns.
    2. Determination of Measurements – define the points of measure and landmarks that are necessary to properly alter a pattern.
    3. CAD Program – Use a CAD program that has all the necessary tools and functions to work with the required measurements
    4. Business Process – Have available all the parts to automate, error check and process the orders in an efficient way. There has to be a business case for automating – yes! So this means a marketing and business plan that requires this technology to meet volume and business goals.
    As with some companies in the past such as Levi’s, one of the success factors above, failed. I suspect it was to do with correctly defining the measurements required to successfully alter a pair of pants. Using pattern making, alteration and grading methodologies as learned manually are logically programmed to the pattern. In the case pants, the waist is not the waist. It is the top opening of the pant. If the rise is related to true waist, many of us would be wearing our pants under our busts! The rise is actually preferred rise. It is about comfort and style. So for MTM measure systems there needs to be way to input preferred outseam and inseam and maybe pitch. Measurement Extraction can be programmed to capture girth measurements at these expected preferred spots. Measurement extraction can also capture front and back girths, seat prominence, etc to gather measurements to alter a pattern with the rise at the preferred height and with the right amount in the seat.

    There is also a point to be made about how much 2D, 3D and scanning technologies have improved in the past 5 years. We all know that as soon as we buy a computer, there is a better one on the market next week. Software does not evolve as quickly, but its safe to say, the 3D scanners used by Levi’s have improved. So take the draft methods and formulas you used to draft a pattern manually and program it to the pattern. All the measurements that you used to draft can be extracted from the scan. All the technology, skills, experience are there, its just a matter of gluing it together in a way that is unique to your product.

    Future Developments for Consumers:
    1. Web deployed realtime draping of MTM on a model morphed to your measurements. You can see immediately what the style looks like on your body. Why? Because its part of the shopping experience and it helps to merchandise an outfit specially designed for you.
    2. Scanners that send your measurements to an online database or an ID card. You slide your card through a store scanner and the system recommends sizes and style for your body type and measurements
    3. 3D Designing – Design your style in 3D and then have it generate a 2D pattern. Many other industries are already doing this (upholstery). Apparel with its varying fabric and stitch characteristics has made this more challenging. But great strides are being made in this area. Designer have been demanding this for a long time!

    So keep your eyes and ears open, 2D to 3D and scanning technologies are not going away. Start learning about it now. You never know, your next job may be stitching patterns and emailing your virtual styles to China!

  8. 3D pattern CAD software

    In June I’d written a post on 3D body scanning that generated some interesting comments. While I still don’t think body scanning will become a viable option for DEs anytime soon, the application of 3D modeling in pattern making software…

  9. Dave K says:

    3D models greatest advantage would lay in the ability to resize the form and create a redrafted pattern (I assume reverse “engineering” capability). Reference points for any design/image printed could have reference points put on the resized form and they’d now be properly placed in the flat pattern in 2D. Reference points are easier than the actual image). While this could be done with a 2D CAD system, as many discussions have brought out, as body measurements/types change so does how the garment will lay/fit on a person. Example, as bust size increases/decreases pockets or the designs may need to be adjusted for “strategic” reasons. A 3D model would allow you to see it immediately as it would be worn.

    I’m curious does anyone use the off-the-shelf POSER software (about $200) and then draw their concepts onto a body using say a WACOM digitizer?

    Questions and concerns regarding the hardware key, based on personal experience:
    • Can you view and print files without it?
    • Can the files be stored in non-proprietary formats, at least the 2D files, so if the key should become lost/damaged all of your work isn’t lost/locked out also.
    • Companies do get bought and sold, go out of business or discontinue product lines so a replacement key may not be available in a few years. Note that was the case in the product discussed where one company split into two, who holds the IP rights can be a big issue. How happens to your license, etc.

    Dave

  10. Aykut says:

    The best value lying in this application is to provide 3D Data for virtual try-on. So that the customers will not have to wear on and off the garments – because it is time consuming and annoying – especially for men. Also customers will upload their data to the web and try the garments online to see how the garments look on them. But for this to be available in markets very soon, the 3D Scanning devices should be economically feasible for the boutique shops to buy them. It’s a nice invention. It can be used for hats and shoes as well. Or other accessories.. There is no need to loose time for doing a complete manufacturing package. Visual appearance softwares should be developed quickly.

Leave a Reply

You have to agree to the comment policy.