Inclusive Data Collection in the Apparel Industry | Humanetics | PI Apparel
Humanetics has spent a decade collecting 3D body scans across four continents to build inclusive size data. They're expanding beyond standard measurements to include wheelchair users, amputees, and people with Down syndrome, using new floor-based scanners that don't require static poses. The i-Size portal now powers dynamic size tables for apparel brands.
Chapters
Thank you all for coming and your interest in my talk. First of all, I also want to thank the entire PI team for the great event at this amazing location. And before we jump into my presentation about the inclusive data collection, I have one remark. It's probably my 10th year at PI conferences, and it's the third time I show up with a different company name, though I always represent the same company. So we used to be Human Solutions, Solutions. We changed our name, I would say, 5 years ago to Avalution, and 2 years ago, 1.5 years ago, we have been acquired by a technology company out of Detroit, Michigan, Humanetics. And now in October, we finally also, yeah, got their name. So now I'm representing Humanetics, and I just hope it's gonna stay with this name for a couple of years so we don't have to always explain the story how and why. However, I'm really glad we are in the Humanetics Group right now because it's a great industry and a great industry leader in terms of body data, so we can really take advantage of the infrastructure, of the resources that we have available right now. No matter what name we used in the past, we did quite an extensive journey in regards of size surveys and data collection. We started that in 2009, 2010 with our size Germany study, and since then we expanded our portfolio in this regards by doing another size survey in Italy. We did one in the US and Canada, size North America, and we are right now doing one in China, hopefully being concluded by the end of the year. We got a little bit of a setback during COVID but we are now in a very good way. We are in the last quarter of data collection, and I guess we're gonna be ready soon for the data processing. Um, the— our goal is always in these data collection and site surveys to be representative. So we always measure and scan a representative subsample of people to then provide the industry with valid, statistically valid data they can rely on, and they know exactly how the population in these regions or countries look like. And we always did it with 3D technology, so we always used 3D scanners measuring people in 3 to 4 positions, and then we took this 3D data to extract the measurements from. In addition, we always collect pretty large extensive demographic data through to a questionnaire which supplements then the data we make available for the industry. Over the years we saw, okay, full body scans is one thing, but the industry always demanded more. The requirements increased, so we started also to do more like customized studies, extended studies, scanning hands, scanning feet, doing a measurement campaign on head and face data. So we saw, you know, everything is moving forward. So we also changed our processes and expanded our portfolio in this regards. When we will conclude China, the journey certainly doesn't stop there, doesn't end there. We already talked even today with some people about the next measurement campaigns we may do in India, Southeast Asia, Japan, South Africa. These are all countries and regions we hear the demand from the industry, and certainly we want to provide data to them as well. What is unique on our end is, I said earlier, we do statistically valid data data samples, and therefore we also want to ensure the industry, if they are using this data, they compare apples to apples. And therefore we always implement our data into our global body data portal called i-Size, which allows the industry to analyze, to slice and dice the data we are providing. And as I said, very important, the measurement definitions are always the same, so they can rely on— if they compare the chest girth from the US to the chest in Europe, that they are looking at the same measurement and the same measurement definition and the same way a measurement has been taken. In this iSIZE portal, you not only have access to the data and you can read it. I mean, back in the days, very often, if you got anthropometric data from somewhere, you got a PDF or you got a book and you got percentiles and so on. This is not the case in this portal. You have a lot of analysis options. You can— you have access to the statistics of the measurement. You have access to measurement correlations. You can check your size table. You can really take a look on different body shapes, and you can start doing a couple of different sizing strategies. What may fit on this market, and maybe right now it doesn't fit on that market, so you can do adjustments. And in the end, you have a very good idea of your potential market shares in certain regions. And it's really very easily accessible by our industry partners. To a web portal. So that's just a little bit of an outline, what we did, what we are still doing. However, my presentation today is about, you know, being more inclusive. What we found, and again, it's also— I don't present here a ready-to-go solution, but certainly over the years we realized and we have seen, okay, what we are usually our data collection is very often the regular consumer. We didn't take a lot of thoughts into being inclusive and catering also, you know, groups of people that may need a little bit more attention. And they are also very, very important for the industry. And this year, what you see right now on that slide is the official definition of inclusivity. And it's, again, For me, also very important that, yeah, that we are providing an equal— in our case, not the equal access, but the equal data also for these groups that are otherwise excluded or marginalized, such as those having physical or intellectual disabilities or belonging to other minority groups. And as I said, that's probably for many of you, it's nothing new. But we as data collectors, as well as our industry partners, It was at some point something new. I remember when we started our SAIS North America survey in 2017, we talked to industry, we talked to our sponsors, we talked to our collaborators, and the question came up, um, are you also collecting, um, other genders than male and female? And back then, at this time, we didn't even think about that. Certainly implemented it. We changed our process in our questionnaire. We added a third category plus a field where somebody can put in their gender identification. And in the end, it turned out it was quite a large group we didn't probably cover before if we wouldn't have this question implemented. So In this case, we saw, okay, we need to change, we need to be more inclusive. And therefore, we thought about, okay, who is right now not covered in our surveys? Here we have a couple of examples of people that are currently definitely underrepresented in our surveys, probably in general in data collection. Campaigns in general. So for example, seated individuals, little people, individuals with amputation, limb loss, Down syndrome. So again, this list here is not— it's not a complete list, but it's a starting point for us where we see also if we talk to our scientific partners as well as if we talk to our industry partners these are groups, they should be included, and there is also from the industry a demand to have data available for these groups to cater them better in the end and give them better products. Because, as I said, what are the challenges for the industry? And this is not only limited to the apparel industry, by the way. This is in general for the ergonomics, for aviation, for automobile, Today, the mass market industry really has a couple of challenges because there might be a narrow view of the disability. What does it mean? What does it mean for the people? There might be design decisions which are not evidence-based. That means they probably think, okay, this could be the right way to change the design of a garment. But who knows, is it really? And in general, there might be a limited knowledge of the target population and on their needs. And therefore, in our opinion, on the one hand side, there is a missed opportunity for the industry to capture the market in general, also, I mean, in the end commercially. But on the other hand side, certainly a missed opportunity to, yeah, make the life for these groups easier and better. Including these groups, and if you think about, I mean, I've just shown a couple of examples. There are certainly many, many more. And also, I don't want to limit it to people living with disabilities. It's also other minority groups that might be underrepresented. So it needs to be really A company has to put— or we has to put some thoughts together. How do we define a population in future measurement campaigns? Do we want to make a special measurement campaign for these groups? Do we want to include them into a regular size survey for a country? So in the end, we want to make sure, as I said earlier, we are statistically representative. The challenge here is, can this be achieved with, let's say, having a statistically valid data pool for wheelchair user? We don't know yet. We have to think about it. We are partnering with universities, especially in the U.S., thinking about that, how we define a population. In the end, it could be also measuring or scanning individuals. Individuals and providing the data more fit model-like, so that we don't come up with the, you know, statistically valid data pool of wheelchair users between 18 and 75, but we have data available that is helping the industry to make in the end products better. Certainly it also needs a change in our survey location selection. If I think about, I refer Again, back to the SAIS North America study, we had a great scanning truck. This truck, you know, drove to the entire states and it was a really nice mobile tool to go to the people rather than asking them to come to us. But if I think about today, this truck had one staircase to enter and to exit, that means a lot of people, for example wheelchair users or amputees or something, they weren't even able to get in there. Or locations in the third floor of a university without an elevator. So these kind of things certainly needs to be, or need to be considered as well. And as I said, it's a change of mind, it's a process, no ready-to-go concept yet because there are So there are many, many different things that need to be considered. Same with the personnel. The beauty, the advantage of using body scanners for these measurement campaigns is you don't need very, very skilled people on site because all they have to do is guiding the people through the scanning process and all the anthropometric skill set is required if you do the measurement extraction based on the scan. Now including people living with disabilities may require a different process, a different skill level of personnel as well. They have to be understanding, they have to be catering, they have to tell people exactly what's going to happen. So this is also a change of process, the change in the approach how to do that. And certainly in the end, I talked about earlier about our questionnaire. We want to also allow those people to have their voice being heard. So we want to change our questionnaire so they can also say, hey, what are we missing right now? Where are the issues? Where are the problems we are having in our daily life? So we can again take this demographic data complementary to the measurement data and make it available for the analysis. And the most important part in the end, as I said, we are a tech company, we are a 3D company, so we are always using body scanners. So the question now is, okay, do we have to change our equipment? What we always did is on the left-hand side, we used our laser-based scanner, and the scanner has a platform. So a wheelchair user cannot get on this platform, amputee may not get on this platform. Other people may have difficulties stepping up this platform. So we have to think about what are the equipment that— what is the equipment that we can use? Is it for some people, do we have to go back to the traditional measurement tapes? Which I guess we have to, because there are occlusions in the software if you skin somebody, in a wheelchair. And the good thing is there are also alternatives. There are meanwhile scanners out there, they don't have a platform anymore, so you can walk on the floor into a scanning device, which makes it much easier. You see on the, on the lower button, that's a picture of the Move4D from our friends from IBV in Spain. It got a very, very nice and large scan area, and it's also on the floor, no platform, nothing is needed. A big advantage here, especially talking about the 4D scanner, is also you are not limited to static scanning. The laser scanner we used in the past usually takes 10 to 12 seconds to scan a person in a static pose, and very often we ask them also to stand in a reaching pose or in a seated pose with the arms angled and so on. But certainly people need to be able to hold that pose for these 12 seconds. Dynamic scanning certainly is very— the big advantage, you can do all of that now in a dynamic way. So you don't have to ask somebody to hold a certain position, you just ask them to get into certain positions, maybe even in kind of a flow. So this is very important. As I said earlier, no ready-to-go concept yet, but there are— the good information, the good news is there are alternatives, and we are evaluating these alternatives. And most likely in one of the future studies, we're going to use advanced equipment, advanced processes for these kind of data collections. When I said earlier the surveys we did, we always had a very, very advanced QA or QS process during the scanning session as well as after the scanning session. I said earlier, the people on site usually they don't need to be highly, highly skilled, but during the measurement extraction, there we need the experts. Experts defining certain landmarks on the scan. And here, when we now talk about including some of those groups I mentioned earlier, there might be an additional level of knowledge available. Here, as I said earlier, we are talking with universities. We have access to some experts in anthropometry, especially for people with living disabilities, and that means we we may implement a new and advanced process in measurement extraction, in quality control, and there will be definitely also new measurements required. I said earlier, we always go by ISO 8559, ISO 7250, and ASTM. We may need to implement new measurements. Also here, we have to talk to the industry, to science, to research, to come up with a new or advanced measurement set. The data organization in the end will be pretty much the same as we always do it. We implement the data into our database. There is a third level of quality control to identify potential outliers or mismatches, demographic versus measurement, and there will be done correlation checks. A very big topic in my opinion. We also talked today to a couple of interested parties here on the event about avatars. As we all know, most of the avatars are totally perfect generic avatars, yeah, which work perfectly in all of the 3D tools. However, we did already some testing a couple of years ago, or not years, maybe a month ago, creating an avatar for a girl with Down syndrome. You see that here on the screen. In addition, we are right now in preparation on creating a mannequin, a physical dress form, especially for that girl. And our partner, that is an image from the University of Delaware, we found out that the regular mass market garments, they don't fit on her in this case. So she changed the design. She got a better design and it fits much better. So therefore, also the avatar creation in the end is something that we are considering. And here in the end, the circle closes. As I said, even for these advanced data collections and more inclusive data collections, we always plan to put that into our eyesize, make it available for the industry. Right now we are working on a relaunch of eyesize, so there will be a new version available in April with a better performance and it will be much easier to use. And what we already implemented are new features new functionalities that allow for a more in-depth analysis also for these inclusive studies that we may do in the future. And it's also ready for new ways of data collection in regards of— I said earlier, 4D scanning. So there could be a dynamic size table rather than a static size table, meaning you get a range of measurement, somebody can take instead of one static measurement. And we also want to include additional metadata. I said earlier, back in the days, we got the demands for hand, feet, heads. Now we get more demands, why don't you also collect voices? Why don't you collect mimic or gestures during these measurement campaigns? And this, again, is complementary then to the measuring effects, and this is something that we really want to provide to the industry. And with that, I hope I were able to give you a little bit of an insight about our approach, how also our thinking changed over the years, as well as the challenges that come with that. And I'm looking forward to see you on our booth. We can talk more in detail, and I'm happy to take some questions.
