Revolutionizing Apparel Sizing with Data and Technology | Humanetics | PI Apparel

Welcome everyone. I'm really excited to be here with 3 wonderful guests and maybe, you know, you go and introduce yourself.

Yes, nice to meet you and very happy to be here. My name is Anke from Humanetics Digital Germany and yes, I'm really happy to join my colleagues here in the panel discussion.

Hi, I'm Lynn Barady from Oklahoma State University. I am thrilled to be here today.

With Humanetics.

And likewise, good afternoon, everyone. I'm Rory Millam from Figure Forms in Cape Town, South Africa.

Yeah, great to have you here. So thanks very much. And, you know, I want to start this a little bit with a, with a story. You know, you might wonder what this has to do with the topic, but I would come to it. So during World War II, the US Navy was trying to figure out whether they can put armor on planes to protect those planes better so that when they go in battle, they come back home. So they looked at every plane that returned from a fight and they recorded all the bullet holes on the plane. And they reached out to the statistic research group. It was like the Manhattan Project, just they did equation and not explosives. And they were top secret installation and a lot of brilliant guys. So they gave them all the recorded ordered bullet holes, and they were saying, well, the Navy thought we will reinforce the planes where they got shot a lot, and we will reduce the armor where they got not shot. Only one guy at this SRG, the Statistical Research Group, thought, nope, we put the armor where the planes that return have no damage, because all the planes that are damaged at these areas may go down and not make it back. So they started with the wrong assumption, thinking that they have a solution. We just, you know, reinforce these red holes. But actually the critical holes are in the green area. So that made me thinking that very often in personal as in business life, we might draw some conclusions on, you know, insufficient data. And that's why we are here to talk a little bit about, you know, the power of data and how data, the right data can help to drive decisions. In, you know, size table development, avatar, physical, digital, you know, fit forms, etc. So after my little story, Lynn, you know, you're not just a sizing and fitting expert, but you also have, you know, tremendous knowledge about history. Tell us a little bit more about the history of the sizing.

Sure.

Absolutely. For women, They haven't had sizes for very long. The first sizing study for women happened in 1941, where two women were charged with creating a study of about— they measured about 15,000 women, but they threw out a lot of data, particularly those of people of color. And they ended up with 10,000 white women. At the time, what happened is that— the industry was losing money. As a matter of fact, a Time article said that they were losing $10 million a year due to poor fit. I can only imagine now that would be a great number. They used their— they didn't create sizes. They just did the fit, the body shape. So they reported on the body shapes and the proportions. That was in effect until 1958, when the Mail Order Association of the United States went to— the government and asked the Commercial Standards Department to please conduct another study. We want sizes this time. So what they did is they went back, they reanalyzed the sizing of the 1941 study, updated some of those measurements from military studies, and created about 27 different sizes. So it went from 8 to 38, tall, regular, and short, as well as if you had full hips, or smaller hips. That standard remained in place until 1983 when all the commercial standards were withdrawn. In 1995, ASTM— I'm sure we've heard of ASTM— they picked up the sizing and they developed sizing tables. Then the next time we really see a big difference is in 2003 when SizeUSA does one of the first huge studies on body data. But one of the problems with— and you can tell this when I tell you that back in 1940s, a size 32 bust was a size 8. That's now a size 0. So the sizes of the humans keep changing, and yet we haven't kept up with that through information. And so we recommended through the research is that about every 5 or 6 years the population changes. And that's because of nutrition, eating habits, ethnicities, and age. So we need to keep up with that. Also understand that sizing standards have never been required. They're completely voluntary. So when we see the vanity sizing, when I fit in a size medium in one but a 2X in the other, or when I travel to foreign country. Sometimes I'm a 3X, which is really depressing. Um, it's not standardized. So that's where we stand right now.

What about the men's? Is there any better?

Well, men's size— okay, so women's are random numbers. Men's are at least with some sort of measurements. But that— it didn't start that way. So we've had some sort of like a small, medium, large in men's since before the 1800s, and most of that was for, like, so— soldiers and sailors. As a matter of fact, we had the first arsenal for the military in Philadelphia in 1800, producing 2,000 to 3,000 suits per week. They weren't great fit. The next big time we see any sort of sizing study is in the Civil War, where they took measurements of men during the war for— Physical health. But we have to remember that women were able to pass the physical health exam and get into the military. So they weren't very good measurements. It's not something that I would certainly rely on. And in the Civil War, they created the sizes of 1, 2, 3, and 4, which was size 36 chest, 38, 40, and 42 chest. And it's not until the end of the war that we start seeing sizes 5, 6, and 7. All data says it was poor fit, incorrect sizing due to the poor manufacturing, and the soldiers just took what they could get. So years after the war, two more studies come out sort of saying that this— these data is not good, but they really didn't do much more until World War I. So in 1920, when the military was disbanding World War I soldiers, and this is the first time that measurements were taken specifically for sizing of uniforms. They measured over 100,000 men and came up with 22 sizes of shirts and 24 for pants. Again, all white men, all military age. But the military at that point went from a 30% of uniforms returned due to poor fit to only 1% within 7 years. So it worked.

Yeah, great. So Lynn, you mentioned, you know, size surveys and, you know, measuring people. Anke, you know, what are size surveys? Give us a little update about this.

Yes, I think the latest size surveys in the US have been done in between 2017 and 2019, where about 18,000 People here in the US between 6 and 75+ years have been measured. I think we selected over 50 locations in the US to cover very well the complete statistics, social demographics that need to be covered because this is very important. In parallel, we did it with 3D body scanners that allowed us to capture completely 3D shape of each of the subjects in different postures. And I think what's quite important, we always do really a combined social demographic questionnaire where people are asking basic questions about age and gender, such things, but also about wearing preferences, ordered sizes, preferred brands. And this allows us to correlate data from such things. In addition, we took about 4,000 foot scans also. Sometimes we also scan hands just to have a real big variance of data that is available there. And why is this so important? I think as Lynn already said, over a period of years, the complete distribution of the population changes. We want to see how they change. I think 5 to 6 years is a very ambitious goal to fulfill because also the industry needs to have a certain time to implement the data. But about 10 years is a very good period to see if something has changed.

Yeah.

And I think Lynn also mentioned that, you know, there were some sampling errors in the beginning, you know, white, predominantly white women. What did you do with SAIS North America to prevent this? To happen?

What we are doing, we are starting always based on social demographic data from the ministries, statistical data to see the distribution of people in different ages, different genders, different regions. Then we are checking which group do we need to sample separately. And based off that, a complete sampling process has been set up and the plan And if we are then changing the measurement sites, each of the scan that has been done is checked directly according to those sampling parameters. So, and if we see we already have enough white women, we can put this information or give this information to the measurement site directly to say, okay, here we are already fine, but we maybe need children, what is always a big challenging topic between 6 and 10. And so with that, we guarantee that on the one hand, we get a real big variance, statistical valid data, and on the other hand, that we do not carry up too much scans that we later on do not need.

Okay. And you also said you use, you know, 3D body scanning.

Yes. Oops.

Oh, I see. Sorry about this. You use 3D body scanning. So what benefits do you get from a 3D scanner?

I think what's first of all, it's really important that we have a very quick technology so that movements during scanning process do not have any influence because it happens so quickly. And we have a really high precision with 1 millimeter so that there is not a lot of variance coming from the technology scanning. And at the same time, on the one hand, we can implement a lot of automatically points of measures into our software according to ASTM or ISO measurements. But at the same time, if at any stage later on you need additional measurements, specific ones for specific projects, you can do that later on because you have the scans, you can remeasure them. And so based on SISE North America, as an example, we had a lot of remeasuring according to specific head data. Or hand data later on.

Yeah, I remember during COVID all the measurements for the—

Yes, for creating the mask. Yeah.

And if I can add to that, Anke, I mean, essentially companies are producing products that are being designed to be worn on real bodies, you know, with real body shapes and anatomical attributes. And the tools being used for that development and fit should be representative of that. Not a simplified, amorphous or, you know, unpostured body, but something that will correctly show the fit and the balance of the products being designed for it.

Yes. And maybe if you look at the— if you had a look at the big range of body shapes, this is something that you just can cover with 3D technology.

Yeah, yeah, yeah.

From my understanding, Humanetics Digital just introduced with some partners a whole set of new offerings like size charts, fit forms, digital avatars. How can those be used and make daily life easier?

Well, we've been having great fun with it, actually. So we use a lot of the 3D data to teach our students about sizing and about sizing in a population. We also have your 3D body scanner, and we scan the models used for our design students. So in their senior year, our design students create a whole line of clothing, and so they bring their models in, they get them 3D body scanned, and we bring them into our software to create digital models, and the students create their garments in the 2D environment— I mean, the 3D environment first. They can check for style, they can check for fit, for fit. They can tweak things. And it's not until they've been signed off that the fit is good in the software that they then print out the pattern and make up the first muslin. And then they can go right into their fashion fabric. And that has cut it down from, first of all, crying students because they're on their fourth muslin, having to call their model in once again try this garment. Our students have streamlined the process. They're making more garments because of it, and they're so much happier. They love the process. And I think that's been the best, best part of it.

Interesting. Other question, you know, talking about avatars and, you know, what makes Size A avatars so special?

Yes, I would say First of all, we did a survey in the industry in the US and Europe last year, and we wanted to know about the requirements of the industry on avatars. What is really needed? What is important? Which priority has which kind of feature? And I, first of all, I want to point out that our survey showed that only one-third uses avatars from the 3D tools because they already deliver an option to integrate body measurements and get avatars out of it. Another third developed individual avatars. We also asked about why, because very often the database that is available in the tools is not that covering all the shapes that we've seen here that we have been collecting. And this leads maybe in some yes, possible not so nice shapes. The relations are not that right. And so the expectation of the industry people is that really those integrated 3D data lead to exactly the shapes that you see in the population. So first of all, this is very important. And I would see— I would say it would really help very much to use the base of these most updated survey data as a base also for avatar generation, also for individual ones as well as for the creation within the tools. And I would say therefore we on the one hand used a standard size chart that best matches our data cloud from Size Age. So we created a size chart with different figure types, 8 groups there. And we used the measurements to create avatars according to this. And as in the background, we still have those 3D data from the size surveys, and we average 3D data as representatives of sizes, this on the one hand gets a very natural appearance. So some specific body characteristics are kept. They can be seen on the avatars if proportions change or if posture is changing a bit. And on the other hand, we create complete size sets that with that have a very consistent behavior in between. So this is the background for these new developments. And yes, they have several features, measurement lines to exactly show where the measurements are. So if you get your garments in the transparent mode, you can very easily see them.

Yeah. Other question. How do you create an accurate physical fit form from 3D avatars?

Well, first of all, we couldn't do it without your valuable data. I think one of the critical first steps is, again, to look at that and validate those base avatars, that they are consistent and accurate with the data, the measurements. The posture that you're trying to achieve, and the, the body shape corresponds to those target measurements accurately. Quite importantly, this needs to be done across an entire size set, not individually and having to go back at later stages to, to adapt everything. So right from the get-go, you check that consistency across the entire size set. The— from those base avatars, we then will model a virtual fit form prototype. So that is— it's CAD engineered from the base avatar, and we start incorporating the essential design features of what that physical end product is going to have. So for example, features such as the removable arms, the attachable head, the arm and leg positions, adaptations for crotch shaping and the intergluteal cleft, even fine modeling of anatomical attributes, muscles, breast root shaping that are essentially critical to, to fit at the end of the day. That all gets done as a fine-tuning process on the virtual prototype. We do additional CAD modeling. To design more advanced features like your articulated elbows, or maybe in the case of babies and infants, our products have got articulated shoulders and hips and knees. And then finally, these— well, not finally, but the next step is these 3D CAD files are used directly for the next additive manufacturing stage. So we implement a 3D printing process to generate a physical prototype. And for those who've got any experience with 3D printing, it's not a simple process of just sending the file to the printer, but certainly needing to control for the print size, the, the materials that are being used, the environment, the influences that has on a successful end print is is important, and that's something we've spent several years refining and perfecting. We then take those 3D-printed components and we will do the toolmaking directly from that. Um, the production molds are created directly on those 3D-printed components. So when you open the molds, you get the negative cavities in which the products are produced. And this is quite important because Due to the features and the finishes that the forms have, essentially they need to be made in— this will dictate the materials in which they are produced. And the material processing means that just about everything needs to be produced from a mold. We are doing some more advanced applications now with production 3D printing, but the importance of having the mold is, at the end of the day, to create the consistency that every product of a particular model will be identical to the original.

Yeah. So we talked about this, or you talked about this a little bit. So just maybe summarize again the importance of having a digital avatar and going from the digital avatar to the virtual avatar— sorry, to the physical fit form. Sorry.

Yeah, maybe I can answer that a bit. Our avatar survey also showed that nearly none of the companies rely currently only on virtual fittings. So the reasons for it are, yes, are numerous, but 90% of the customers that we were asking used a combination of avatar and fitform at different stages. Of the product development process. And I think if you do so, if you use this combination, the consistency between the avatar and the fit form is really very important. So if you want to see the same situation virtually as physically, they need to be consistent. And we have developed a huge process where we constantly compare shapings in the development process between the avatar and the physical form to really guarantee that the adaptations that need to be done due for functionality aspects, that they do not influence the shape that much. And so this is really very important. And I think everybody here knows due to cost aspects, very often, The fit forms are only created in the sample size. Fittings are only done in the sample size. Some companies only have models in those sample sizes because it's always hard to find them. And I think what's a very good and valid combination is to have the sample size as avatar and fit form, but having a complete size range of avatars where you can check the grading on and see, because bodies do not— proportionally develop. They develop in different body types and different size ranges. And so it's really very, very, very important to, on the one hand, have this combination of avatar and fit form, but also use checks at a wider size range. Yeah. I think only after the approval really of a— on a physical fit form, The most of the companies rely on the results and say, okay, now we are fine with that. And yes, I think this virtual technology already helps us very much in the development process to avoid mistakes, to avoid wrong gradings, to, yes, search out collections or to bring out collections that perfectly fit in the sample size, but not over all the sizes. And here we are at a very good combination of avatar and fit form, I would say.

Yeah, I agree with that. And to even, well, take that back a step and to reinforce it, you know, we've been working with companies for over 35 years and we've seen that, you know, that typical garment approval process, as you mentioned, it's done on that one, maybe two, base sizes, but it's typically done doing comparative fit trials between the physical form and the live model. And although that is beneficial, you know, the live models can provide valuable feedback on the actual fit and the feel of the garment, the movement and the ease. But it can be exceptionally time-consuming, very costly. And not always the most accurate. You know, a lot of companies aren't able to find fit models that perfectly match their body fit standards or measurements, and those fit models are not always available. They can, they can change in measurements over time. Certainly for factories and suppliers, they are not always available. So looking at that traditional sort of process that should really be brought back into the more modern technologies, comparing the digital with the physical forms.

Just to wrap this up, also quick question, is every Fitform the same or are there differences?

No. Well, they shouldn't be. Let me put it that way. So I think from our perspective, adaptability and innovation is key. In the approach to solving certain customer problems and being able to produce some unique products. I mean, this is an example here of a particular fit form that we developed for a wetsuit and sporting apparel company, and they needed to be able to test a certain range of, of movement during the product development phase. We've taken this kind of an innovative approach to designing many specialist features for different customers on our forms. And we've only really been able to do this more efficiently with the power of 3D and digital tools and technology at our disposal today. But it also comes down to what customers are used to working with. I think a lot of perception in different markets is all the traditional fitting forms in use. Customers are used to certain features and functionality on them. I think they should really look at exploring different opportunities or different features that could actually add to their workflow and productivity.

Great. Just, you know, to summarize this, I think we're running out of time as well, but I hope we were able to show you a little bit, you know, importance of, you know, to have a right data source, you know, SciSurvey data. And then all these steps you can take to go from, you know, the right data measurements, socio-demographic information, go to, you know, physical— sorry, digital avatars, go to physical fit forms, analyze your size shots before. So it's really important to have the right data to start all your developments with. So to do this, I think we have maybe a minute left for questions, but you can also see, you know, Humanetics and partners at the booth number 1 here, just around the corner. Questions? One minute. Good. Because the other guys are getting ready as well, so appreciate your time. Thank you very much. Thank you for joining us here and have a great day. Okay, thanks.

Thank you.