Video Transcript

Deakin: AI for sustainable fashion

I have lots of concerns about the sustainability of most things we do on earth like where it's sort of talking about consumption and production of goods where we're contributing to lots of emissions and climate change and at the same time, we've got all this waste building up that we don't know what to do with, I want and we need to do something about it

In early high school a lot of the teachers didn't believe that I would get a good high school score or ATAR score because I said I'm not trying now.

I'll just try in year eleven and twelve because that's what matters and then I got an alright score but I wasn't like the brainiac, I was just sort of your average Joe that was a bit interested in maths

This is my desk from the work that I did in my PhD. I’ll have to try to remember the title of it. I think it was circular textiles, redesigning textile fibres for a circular economy.

Before my PhD I had no idea about textiles and their impact. There are different stats out there that people nowadays are saying it's the second biggest polluter in the world.

In this experiment today it's a bit of a two for the price of one. First, I'm going to use a machine learning model to predict the cellulose solubility or cotton solubility in our solvents working with some people from A2 I2.

I think it's the Applied Artificial Intelligence Institute here at Deakin. So I’ll predict the solubility of high DP cotton then this should predict we’re going to get 8.05 weight percent of cellulose in there and then we're going to go test it experimentally in the lab, myself and Nilupul. Nilupul is a PhD student.

We don't know how it works because this is quite different material and new combination.

We’ll see.

His project is about separating elastane or spandex from cotton which is super important because just a little bit of spandex can make it really hard to recycle waste cotton textiles so we're going to use waste cotton as the cellulose source and we're going to use ionic liquids as the solvents and we'll basically just heat them up and stir them for about four hours.

Oh it doesn't come off easy that's for sure. Come in Nilupul.

These experiments today will show either high DP cellulose dissolves at this concentration or it doesn't.

If it doesn't we still don't know if it's the spandex or the cotton, but if it does it's because of the spandex in the solution.

Then you can go back to the drawing board and figure out how to optimise that separation which then we can test later.

What we've got here is two, four, six, eight, ten. So they're the concentrations of cellulose  that we're going to put in there, get some Kim wipes from Kim’s desk.

Do you want a go at mixing that? I’m hanging for a coffee. Usually, I have two instant coffees before I start work.

I hope this works. For me it's a win-win because I'll have something to put into the machine learning model. For Nilupul if it dissolves four weight percent or above then we can confirm that it's the spandex causing the problem with his dissolution.

If it dissolves under that I suppose it's a it's a half win because whatever concentration it dissolves at you can test your separated cotton at that concentration and that'll give us the conclusion we need and then I think we can put it on the heat.

So it’s eleven thirty, there we go, so that’s mixing now.

To get the stretchability and comfortability we add spandex in the fabric but there's no specific sustainable method in the industry to scaling up the recycling then so my target is to separate the component mainly focussing on mechanical separation and using environmentally friendly chemicals and at the moment we are observing promising results as well.

Let's get a coffee. I’ll go regular oat cappuccino in that one please.

A bit of a bandaid from chipping my knuckle building some outdoor stairs from our deck it’s worth it in the end. I’m a lot safer at work than I am at home I think. 

We've got our solutions. this is quite dissolved. That's basically liquid cotton almost ready for making fibres with. This is two percent of the cotton, this is four percent and if we tip them one flows and one doesn’t.

It’s not always the case if it’s visually dissolved to the eye that it's microscopically dissolved as well so we have to say it looks dissolved and then we confirm dissolution by looking at it in the microscope.

It's just really cool seeing theories and ideas and equations come to life through experiments.

We’ll leave that there. This is good to go, super important one hand one glove. Carry your sample with your gloved hand, so you don't use your glove on the door handle because you don't want to  contaminate the door handles.

We’ll see if we can find some other regions of undissolved stuff in here. here we go. this is our slide with some dust particles and this moving is our cellulose solution on the slide.

From what I can see so far and I'll just scan all the way up I think it's fully dissolved the two weight percent solution was fully dissolved, so next step for you would be to test your cotton sample at two weight percent in the same solvent.

That gives us a good result but not the best result it would have been the best if it dissolved the four weight percent my model won’t be happy though not accurate

If you're filming a day in the life of a scientist and everything worked perfectly it wouldn't be a typical day in the life of a scientist, it would be a a one in a hundred day in the life of a scientist, because that's what we're doing most of the time we’re checking things to make sure they don't work and then we don't have to do them again but we're not necessarily making the big breakthroughs every day.

Even though today was a bit routine in the work that we did we weren't expecting any big results or trying to do anything too ambitious

I try and look at it in the way that you can't have a bad experiment because you always learn something that doesn't work or what not to do next time so I think it's a good way to not be constantly disappointed when you are in science.

I’ll look at the Cape Sorell buoy data and see if there’s a swell showing up and hopefully go for a freezing cold surf tomorrow morning