00:09Europe and the U.S. are throwing out more than 20 million tons of textiles every year
00:15and the problem is growing, but so is the opportunity for change.
00:20These textiles are incredible resources to feed textile-to-textile recycling processes
00:26and create a circular industry. To reclaim these resources, textiles must be sorted by fiber
00:33composition and the demand for recycled content must increase. That's where the fiber sort comes
00:40in. The fiber sort is a revolutionary textile sorting machine that is changing the game.
00:46It sorts finished textile products by fiber type, automatically creating precise recycling feed
00:52stocks and finally providing a practical solution to return textiles to the supply chain.
00:59But we're not quite there yet. We need your help to optimize this technology and increase
01:06the demand for recycled textiles. Are you a collector, sorter, brand or recycler? Then join the front
01:14runners in the textiles revolution and help make the fiber sort a reality. Go to fibersort.eu.
01:39So I'm going to give the floor to Hans Pond and Maurits van den Putt. I'm not sure who's starting
01:44with sharing his presentation.
01:47I'm going to share my screen.
01:50First I'll try to set it up.
01:56Yeah. Presentation mode and then you can go there. Yeah. So go ahead.
02:03All right. So good afternoon, everybody. So my name is Maurits van den Putten. I work at
02:10Valve on Bailing Systems as a project manager. And today I will give a presentation together with
02:16Hans Bonn from Wieland Textiles on fiber sort. So the presentation today for Valvan, I'm going to give
02:25a small introduction who we are, then talk a little bit about the fiber sort machine and also explain how
02:31the technology works. And then the second part will be given by Hans Bonn on smart fiber sorting and the
02:40bigger process of which fiber sort makes a part of. So Valvan Bailing Systems is a Belgian machine builder. We
02:50built
02:50equipment for sorting and baling solutions for applications in the recycling industry, the fiber industry and the
02:57used clothing industry. In Europe, we build a lot of the bigger sorting installations that handle post consumer textiles that
03:07are
03:07are getting sorted into different categories and mostly are getting resold in thrift stores today or secondhand markets abroad.
03:17But more and more, there is a demand coming for textiles that can be recycled. And there is also a
03:25big fraction available in the sorting
03:27installations that has a very low value, but still there is a fiber inside. And that's why we designed a
03:34machine to sort textiles
03:36based on fiber composition, color composition and structure. So the machine is called fiber sort.
03:46It's an automatic sorting machine sorting textiles into different categories. Maybe I can eliminate this.
03:56Yeah. So the machine makes a decision based on fiber composition, color composition and fabric structure,
04:04where it should be sorted, where it should be sorted. The machine exists in different setups, we have automated feeding
04:12systems with robots.
04:14But we have also manual systems where people load the textiles onto a scanning system. The scanning system is always
04:23a part of the
04:25installation to make the installation to make a decision what it is. And then there's a series of blowers which
04:31you saw earlier in the video that blow the textiles based on the three
04:36predictions we make in the system into the right category.
04:43The fiber sort machine here is a 3D layout of the machine. And the production rate of the machine is
04:50more or less one sample every second.
04:52So we sort textiles on a piece per piece basis. And we do that because there is a very big
05:00variation and we need a lot of sorting categories.
05:03And there are also other setups with optical sorting machines, but there is only a selection of one fraction.
05:10But today we have a fiber source and facility of Hans with 90 sorting categories.
05:18So therefore we need to put the samples one per one.
05:23We can process with the scanners 3,600 samples per hour and more or less there are three pieces of
05:31textiles in a kilogram.
05:33So that makes a throughput of 1,200 kilograms per hour.
05:39And then the feeding options as explained, it can be a manual feeding process, but it can also be automatic
05:45with robots.
05:47To be honest, the robots are today a little less fast than the manual operators because we are still optimizing
05:56the setup and fine tuning everything a little bit.
05:59But the manual can be can be very quick.
06:04The color sorting is working on two predictions.
06:08So the first one is we make a distinction between single color and multicolor.
06:12Single color is everything that can be used in a mechanical recycling process where we can eliminate the need for
06:20dyeing.
06:21Multicolor is typically textiles with prints, which you can see in the right picture.
06:28And then for the color sorting, the second step is that we make a very precise prediction of the color.
06:34So, for example, we can sort blue, red and different shades.
06:38And nowadays we sort 15 different colors for certain fiber fractions into all different color shades.
06:47And this is very precise.
06:49It can be more precise than the human eye.
06:52But the more precise we make it, the more categories we get.
06:57So we try to find a good balance between a number of categories and also the volume of colors.
07:08The fiber source technology is based on near-infrared or in short, near.
07:14The scanner emits a near-infrared light that is reflected or absorbed by the scanner.
07:20And the reflection profile of each textile composition is unique.
07:25And we use that information to make a prediction on which fibers are inside the textile.
07:32One scan takes more or less 50 milliseconds.
07:36But to get the pieces of textile piece per piece, we can scan more or less one sample every second.
07:46In this picture, you can see in red a spectrum of pure cotton.
07:52In blue a spectrum of pure polyester and in green a mix between the two.
07:58You can see it's not easy for the human eye to make a link or to understand these spectra.
08:04The spectra represents the amount of reflection of each wavelength of the light.
08:12So what we see is that the same fiber categories, for example, 100% polyester, all the lines are grouped
08:22nicely together.
08:22And then we create a machine learning algorithm that predicts which fiber is on the scanner based on these spectral
08:31information.
08:34So why do we need fiber sorts?
08:36To sort by fiber type with a very high accuracy and increased speed.
08:41The alternative is a manual sorting with a lot of limitations.
08:46A trainer would need more or less six seconds to read the label and then still has to take an
08:51action.
08:52Youth clothing also doesn't always have a label and some of the labels are inaccurate.
08:59So it's not possible to have an accurate sorting based on a manual process without scanning equipment.
09:08All right, that's it for my part, Hans.
09:12Yes, it's interesting to hear you, Maurits, as a young machine builder, because there is no history in this whole
09:19thing at all.
09:20Where does this come from, this fiber sort idea?
09:23It's already, what, 13 years now that we try to develop some kind of technology who is able to add
09:33value and to make worthless material circular.
09:37It started a long, long time ago, and there is a very good productive machine now.
09:46And what happened is when this machine finally got productive, we still had garments.
09:54And the industry was not really interested in garments, the industry wants a raw material.
09:59So we started looking into building a machine that is able to remove all contamination, remove buttons, zippers, but also
10:10labels.
10:11Because almost all labels are made out of polyester.
10:15So it doesn't matter if you have a cotton garment or wool garment or nylon or whatever.
10:20There is always this polyester label, which is a contamination for whatever you would like to do with the material
10:28afterwards.
10:29So what do we do?
10:30We have the garment, we cut up this garment, and with a new technology, which is still under development at
10:37Valvan Baling Systems,
10:39we try to take out all this contamination, the buttons, the zippers, but also the labels.
10:46We call this innovation the trim clean, which is not done yet.
10:50We are still developing that.
10:52But, well, maybe Maurits can tell a bit more about this later.
10:59But this innovation gives us the opportunity to create post-consumer clippings.
11:04We know post-industrial clippings, but we are heading for post-consumer clippings,
11:11which creates a raw material that we can fiberize.
11:16And after these fibers, we are able to do spinning, more or less by adding virgin material.
11:26But we just heard Nicholas, and they are already able to make very acceptable genes with 50% post-consumer
11:35content,
11:35which is quite an achievement.
11:37We call this whole process, including the software, we call it a fiber farm.
11:44And this is, this whole concept is scalable.
11:49So we would like to set up this concept worldwide, wherever there is a large population.
11:55You collect textiles, you sort it, and you trim clean it.
12:00You make post-consumer clippings in order to make new yarns.
12:09Finally, you see that, because I already memorized, it took me about 13 years to set up this fiber sort
12:18idea,
12:19and there was no market.
12:20And what we see now, all of a sudden, is that governments are interfering.
12:26Governments make policy that in 2025, 25% post-consumer recycled content should be in all garments.
12:35That made the industry very restless.
12:38And all of a sudden, because governments are stepping in, you see there is a lot of dynamics.
12:44And now the market is rapidly asking for post-consumer material.
12:52And, well, one big initiative in this whole thing is the Denim Deal,
13:00which should bring us 3 million genes with at least 20% post-consumer content in the next two years.
13:10So, yeah, all together, we are making big steps in cleaning up this textile mess that we are creating
13:20by producing all that textile material.
13:23So far, my part on this presentation.
13:29Maybe to elaborate a little bit on the trim clean that Hans mentioned.
13:35So, the fiber sort can now accurately sort textile bulk material into 90 different categories.
13:44And then we thought we were there and we could start recycling.
13:48But the reality is that to be able to process it on a line to create new yarns,
13:56you have to take out all the zippers, buttons, and labels.
13:59And so now we are working hard with the team to create also a technology
14:03to take out automatically all these zippers, labels, and buttons.
14:10It's still in a development phase.
14:13So, we are not ready yet for the market.
14:16But the first trials look really promising.
14:19And we think or we aim to create a product very similar to post-industrial clippings,
14:26as Hans mentioned.
14:27And it's a good name, the post-consumer clippings.
14:30It captures very well the essence of it.
14:34Right.
14:35And you understand that we put a lot of pressure on Maurits to come up with this trim clean machine
14:41because his machine at the moment, this innovation is our missing link to make our next step.
14:49For now, we do that with zippers, with scissors, to cut out buttons, labels, and zippers.
14:56But, you know, if you want to go to mainstream, we need to make that an industrial solution.
15:03And that's where Maurits comes in with the trim clean.
Comments