Welcome to a new episode of the Stemcognito Interview Series. Today, we have with us Irfan from our Stem Cognito team, and it’s a real pleasure to talk to the second member of our Stem Computer team today in our interview series. So, we already talked to Marta at the beginning of the series. She was the very first person who decided she wanted to get interviewed by me. Now, we have Irfan as the second one. Very brave, well done. Welcome!
Thank you so much. Irfan is a postdoc in the mathematics field at the University Technology Petronas in Malaysia. Just to summarize or start, would you mind summarizing a research project for our audience?
Okay, basically, my research is developing mathematical models for chemical engineering applications. My PhD was focused on developing mathematical models for biodegradable polymers. Biodegradable polymers are used for fertilizer applications. Normally, people use synthetic or industrial polymers which are not biodegradable. After the application, when the nutrients are absorbed by the plants, these synthetic polymers are left in the soil, causing soil degradation. We wanted to come up with something natural that doesn’t harm the soil properties and provides the same duration as synthetic polymers. There was a team at my university working on developing biodegradable polymers, and my role as a mathematician was to develop a model to understand the different reaction and kinetic rates. I would then provide information on how long it would take for the nutrients to be released when using a specific biodegradable material.
Were the polymers naturally degradable, or were there microbes involved?
There was both microbial and enzymatic degradation involved. In the soil, there are enzymes that react with the biodegradable polymers like PLA (polylactic acid) and PLGA (polyglycolic acid). We had to consider three factors: diffusion, enzymatic reaction, and microbial degradation. Diffusion involves the movement of water and nutrients in and out of the soil. For enzymatic and microbial degradation, we combined them into a single term based on experimental data. We determined the reaction rate and how fast degradation would occur when a particular polymer, such as starch or PLA, is inserted into the soil.
Did temperature play a role in the degradation process?
We didn’t consider temperature as a parameter in our study. We didn’t have enough data to validate and incorporate it into our research. Temperature does affect enzyme and microbial activity, but for our fertilizer application study, we assumed a constant temperature.
What did you discover about biodegradation and its dependencies?
Based on different soil types and agricultural applications, such as paddy fields in Malaysia where rice and palm are grown, we recommended a few materials for sustained release over four months. This was crucial for larger-scale applications. In terms of mathematical significance, we incorporated a double diffusion multi-diffusion concept into the modeling. Multi-diffusion refers to the diffusion of water and nutrients, both inside and outside the fertilizer granule. It depends on the specific fertilizer being used.
Going back to your PhD project, was it funded by a fertilizer company? Did you have to show your results to someone before publishing?
Yes, my project was funded by a company in the fertilizer industry. I had to share my results with them before publishing to ensure that the findings aligned with their requirements and objectives.

Great questions! So, let’s go through the text and add correct punctuation without changing words or deleting words:
“Where you’re completely free in your experimental setup? Yes, uh, for that matter, they were. They have given us the freedom, uh, to publish unless and until we give proper, uh, citation and affiliation to the crack. And, uh, there was no, I think, involvement in our results, in terms of modeling is concerned.
As a result, their product is working perfectly. You know what? If your modeling data did not show that? Yeah, but, uh, that didn’t happen. Yeah, okay, but that’s why I’m asking, did this occur to you? Or did this happen? Or did you have to, you know?
There was no, uh, like, we didn’t show it to any of the funder before publishing or before I submit my thesis. It was free from them. Uh, we, whatever results we have got, we submitted. But, uh, I think, uh, because this company which was we funded our project was a government of Malaysia company, not a private sector company, so that’s why maybe which was more free. Yeah, okay, I understand. Okay, so that was your PhD project, which sounds pretty interesting. What about your postdoc project now?
Yes, my postdoc is on a completely different field. Oh boy, oil and gas. Okay, um, right, because they’re also involved in oil and gas degradation. Yes, microbial EOR is a, uh, definitely a very interesting topic. But my project was more into diffusion because as a mathematician, for us, like anything diffusion, we take it as a simple fluid flow, a fixed law, and implement. So, uh, it was, uh, more into diffusion of, like, uh, this project is, we have nano, like, nanoparticle EOR. Is one of what happens is in a reservoir, up to 20 to 40 percent of the oil, only you can take from the well or the oil wells, uh, directly because of the pressure which is there in the reservoir. After that, if you want to take out more, where you pump out the oil forward? Yes, yes, in the natural process, it is called primary or the secondary recovery. We can take only up to 40 percent of the oil, okay? Uh, if, let us say, I have 10 liters of the oil in a reservoir, I can only take four liters. But to take out another six, we have to go into the third stage. That is called enhanced oil recovery, EOR. It is also very famous as EOR, enhanced oil recovery. So what we do is, uh, previously people used to inject hot water, simple hot water. They used to inject, and hot water can change the temperature. Now, here’s the temperature is very important. Oh, it wasn’t perfect.
So the hot water can change the viscosity of the fluid or the oil, and that can help to push the oil and take out as much as possible. And then that one also we can do up to a certain range. So to enhance it more further, we started, people started introducing nanoparticles, polymers. So my this project was on nanoparticles. We had one experimental team, they were doing the co-flooding experiments, they were taking different nanoparticles and they were finding how this nanoparticle will flow.
And then my job was to make a mathematical model. Because they cannot test all the nanoparticles, all the conditions, you know, like pressure conditions, temperature conditions are we talking about, what’s in these nanoparticles and yeah, why do they, how do they have pumping of viscosity as well? What happens is these nanoparticles, uh, if you introduce, uh, along with the electromagnetic waves, oh, okay, so they, because my nanoparticles have the properties of, like, aluminum oxide or zinc oxide, they can react to the electromagnetic waves, yeah. And also the temperature, and they can carry the temperature to a larger point, yeah. And they can also travel further. Uh, what happens in, like, in reservoir geometry, we have many rocks. So there’ll be one area which is covered by the rock, we need someone to go inside and change the temperature so that the liquid can, the fluid can come out, yeah. So there, these nanoparticles were giving a lot better result, uh, and this, like, still, uh, I think, uh, apart from one or two projects, this is not yet implemented in industry. Still at lab scale, people are searching for different nanoparticles. So we use the nanoparticles which will be acting to electromagnetic waves, and we are seeing. My job was to find out the optimum pressure, uh, and what is the concentration of the nanoparticle we needed. Because if you inject most nanoparticles, nanoparticles itself is very costly. Here, uh, when we talk about protection, we have to consider, like, if you are taking out 10 liters extra and you’re spending 10 liters of money on the nanoparticles, then there is no use for the company, right? They require something which is very cheaper and also which can give them enhanced, uh, recovery. So for that, we use different nanoparticles, and then we saw like how they move and how they affect, and this was also a company-funded project by the same, uh, company which I work for. Like, my university is a part of Petronas, oil and gas company, which is Malaysia’s, uh, national oil and gas company. Okay, so they give this project, and we were working for them, and so far, we are in the finishing stage of this plan. That’s good.
Okay, one question, it’s burning inside me, I just cannot comprehend how, what basically you say that you’re doing mathematical modeling, but what does it actually mean? What do you do during the day? Like, do you just, like, start a modeling run test, whatever, in the day, or what? What is your job? What do you actually work at or like, well, work, what do you do?
Okay, my job is to first understand the engineering and the physics. Whatever the model we are building on, like, for example, the nanoparticle flow, so first I have to understand what exactly where the flow is taking place, which is the physics laws representing these flows, yeah? And how they are represented mathematically, okay? So you first need to understand the physical basics, or in the case of the fertilizer, it would be like the biological basics or the chemical basics, and then you can translate this into equations, basically, yes?
Okay, we translate them to equations, and then we implement the boundary conditions, initial conditions, like, initial and boundary condition is like depends on the system of the system we are working on, like, initial condition is something where when the time zero, like, we didn’t start the simulation yet, what was the conditions where each particle was, what is the temperature, everything, and boundary condition is like, if we have a fixed geometry, like, we do or usually simulate in a fixed geometry, like, we define one particular geometry, and we say, what happens at this point three if the particle or fluid comes to this bound, what is the temperature here, or the concentration, or whatever the conditions, and these things we take from the experimental people, we talk to them and we understand what exactly happening. So, our work is to try to mimic as real as possible, and if we can give it up to 90% also, because there will be many approximations, there are many parameters or many things happening which we cannot quantify, yeah? Like, now, in my PhD case, there was, I said, bio, like, they were enzymes as well as the microbial, so microbial degradation, if you want to quantify the different types of microbes in the soil, which react, you know, to different polymers, to quantify it in a very specific way, it is not possible yet, yeah? So, we have to just talk to them and take one approximation and start building our model, yeah? So, yeah, especially in the case of microbes, and as I said, like, the soil doesn’t always have the same type of microbes in there, or I don’t know, as I said, temperature or rain or whatever might influence the whole parameters and their whole breakdown, so is it actually possible to get, like, a real-life model? Or, as I mean, as I said, it’s always like an approximation, but how close do you think you can get to the real life? What is happening in real life?
Not only my research now, like, if I talk about in general, mathematical modeling concept, I think we are now reaching up to 80 to 90% of accuracy, especially now due to the advancement in machine learning, deep learning, image-based modeling techniques, and also a lot of experimental datasets we have and availability. So we are reaching up to 80 to 90%. Not that, uh, I cannot say for some simple applications or simple problems, we can even achieve 95-99%, but if the problem is complex, yeah. And the thing is, even if we achieve 80% accuracy, what is the advantage we are offering? It’s, uh, for you’ve been exactly my next question, yeah. If you can’t reach 100% accuracy, what’s your deal? Why do you do it anyway, just for the fun? I mean, come on!

It is not for the fun, what happens is like, for example, uh, no, no, like, uh, this is, uh, why we do, uh, models. Modeling is done because, uh, experimental study is very expensive, first thing, and if you want to do with 10 types of nanoparticles, you have to bring all those 10 types of nanoparticles and we have to take, but if we can save all that money and you have two nanoparticle did experiment between nanoparticles and we give you for the rest of the nanoparticles, we give an accurate estimation of like, what could be the recovery, how could be nanometer will be flowing that can give the companies, as well as the researchers, an idea what it exactly can happen if they use this particular material or this particular nanoparticle or biodegradable polymers, whatever they are working on, it will give them the idea what exactly this project product will look like at the end, and then if they are social, like, okay, this is very promising, then they can go for that particular specific experiment, so we are saving cost, yeah, and we are also saving, uh, time, manpower, and make sense, it also gives flexibility to go into different materials, see, and go and dive into different things, otherwise, if you’re doing all the experiments, you have to be very specific, you have to take one of the worlds and all these things from the grant body, yeah, so that’s the advantage. So, do you say that like most experiments should be supported by mathematical modeling, or is it just bigger projects or bigger problem problems, or what do you think?
I think in any experimental study, there will be some or the other mathematical modeling involved. If I give you the example, most of the experimental study, you can see everybody plot that regression line, right? And y is equal to that is also an empirical model. Oh, really? I did some mathematical modeling in my PhD as well. That is so good to know. Empirical modeling is, we have experimental data and we fit into some form of equation, like, uh, regression equation, polynomial equation, I know most people use in the experimental people, yeah. And what we do is called as mechanistic modeling. Mechanistic modeling is, we don’t have initial experimental data, we try to understand the physics, use the governing laws of those physics, represent them in three equations and solve it, yeah. So I feel every person will use some, at least some empirical modeling in their studies, yeah, makes sense.
So what do you think, or where in our daily lives do you also use mathematical modeling? Because that is just like the examples we talked about were basically now for scientists, for like biology or chemical studies. But when our daily life do we use mathematical modeling?
We use mathematical modeling everywhere, like, uh, nowadays, even for your ringtones, even for the mobile ringtones, it works on a mathematical modeling principle. Now, more like, uh, the temperature sensors you have in your aircon, you have, like, uh, air conditioner, fridge, all this have certain mathematical modeling small packages. And apart from that, like, now whatever the smart application, if you forget, mobile, laptop, all these things, uh, they basically use many parts of, like, there’s a lot of modeling is going on. Even in this Zoom call which we are talking, there is one feature of the Zoom which says to enhance the beauty, manage the light, enhance the beauty of, like, beauty pictures, you know, on the face. That’s, that is also a mathematical model. That works with, uh, computer vision, image processing, like, we take, uh, images and then we take some features and then we try to enhance it. Okay, now, I think, uh, nowadays, this AI features in the modeling in the photos, especially, will be give you a total different picture, like, the person in the actual and the photos will be completely different, and that works with, and it is possible due to implementation of some or the other form of modeling. Okay, okay, what about the fridge, though, and the aircon? I can’t understand that, like, what does mathematical modeling have to do? What is it doing my fridge? I’m sorry, what? Okay, uh, what I can, I worked with one of, uh, interesting projects on the AC units. It is also very renewable. One of my friends was working on a solar air conditioning system. Uh, his job was to, uh, he was making, using some thermoelectric materials. Thermoelectric materials are like the materials which give extreme temperature at the both ends, like, if one side will be minus 70, other side will be zero, like, uh, one hot, one cold. So you can use based on your application. So one will be very hot, one will be very cold. Uh, there are different types of thermoelectric materials, and he was implementing this thermoelectric material in a house because in Malaysia, we have all the time, we have solar energy.
Okay, and when he discussed this project with me, he wanted to enhance the efficiency of the cooling inside the room. How to do that? There was one way, it’s to, uh, change the temperature speed of the fan based on the outer temperature as well as the room temperature. So we designed one application. Like, if you want to maintain, uh, let us say 27-degree temperature, what could be the temperature of the fan? So that one will be a simple mathematical modeling work will be going on there. The fan speed will change according to the temperatures. If the temperature in the room is going higher, the speed will become more. If the temperature in the room is going low, then the speed… it… it happens with the voltage. It will give signal to the voltage, but all these small studies, we do mathematical modeling, we understand, then only we implement in the real system, and then we go for this. Okay, and even the fridge controller, all these things, they have certain resistance law. Even the simple cathode voltage flow is equal to IRD, say, is a mathematical model. Okay, okay. So, do you think that, I mean, mathematics, obviously, for pupils at school, that’s always like, no, nobody wants to do that. But can you think of, like, this is how you can use mathematics later in your life, and this is why students should stick with this?

Like, how would you explain to a student why mathematics at school is important? This is a very important question for me, actually. When I was studying my bachelor’s and master’s, you know, we used to study a lot of equations, like how to solve these metrics. First year of mathematics, and I was just so glad when it was over, and honestly, I never used it again. Like, all I do, exactly, when I used to be a scientist, I mean, I would still, like, you know, calculate concentrations and stuff, but even for that, I mean, that’s, like, no mathematics that I had at university. So I’m still thinking, why? Why do we need this? The same question was there in my mind when I was studying, and I used to see, why? What is the use of this partial differential equation? What is the use of ordinary? We used to study some of the examples, like bookish examples, you know, to solve, find the area and the curve of the, you know, differential applications and all this. But I feel, if you teach, now, if I teach something, even if I teach a simple number system, I will tell the student or anybody, like, where it will be useful. Okay? And I think that’s the reason many people lose interest in mathematics, especially those people who feel they want to study what is happening in front of their eyes, like the experimental people, like who are more into understanding what’s happening there. So if you keep on giving examples of real life, implement, like, if you’re teaching linear algebra, it’s from school, high school, we start teaching linear algebra, and linear algebra is the backbone of machine learning. Like, if you tell students a simple linear algebra, if you know the linear algebra of your high school and the college, that, I think, will set your machine learning base to go and become a machine learning engineer, artificial intelligence. So we have to give certain real-time examples where it is happening, where exactly we are going to use this, and that can happen when, like, we build a curriculum based on application-oriented and something like we implement what exactly we are doing and start. So that’s my, even my approach, if I become a lecturer, I want to teach, I just don’t want to show them how to solve a particular differential equation or the mathematics. I will tell them, you can do, you know, even for example, like, this is very solid, it may sound very simple, but we can also find out what is the approximate time needed for a milk to boil if you are good in partial differential equation. Okay, like, that is a heat transfer, yeah, it’s a heat transfer, you have the coil or whatever, and there is a thing you have kept, and that is the basic heat conduction equation or the diffusion equation. Whenever we teach partial differential equation, we start with parabolic diffusion. If you tell them, like, you know, this is where you are using this, it’s a simple application here, yeah, then that will build, that people will show interest, okay, I am going to use this, and I think there are a few professors whom I found they have this approach, yeah, of, you know, giving the examples and making mathematics into very practical oriented, and like, this thing, that is, yeah, that’s very true, I guess, yeah, that’s why a lot of people, a lot of students lose interest in mathematics completely. So now we kind of already touched a bit on, like, outreach, and you talked about how you teach your students, and now let’s switch to our common favorite topic, Stem Cognito, that we work together on, um, and you are our data analyst, obviously, for the team, because you’re the only one who has some kind of mathematical understanding.

So my first question for you is, how do you use your mathematical modeling approach or your mathematical thinking for the Stem Cognito project? Okay, the first idea is, now we want to enhance understanding our viewership recommendation system. Like, we are still very new or very young, I can say, but now, yeah, yeah, still few months old, but when we grow bigger, when we have a lot of videos on our platform, we want to understand what exactly the people want. What if you go to YouTube or Netflix? They give you very good recommendations, right? It works with the artificial intelligence algorithm. They know what is your history, they know what exactly you have watched, what you have searched, and that’s based on that, they will recommend you. And we want, we also want to do something similar to our viewers in Stem Cognito. We want to give them, like, if somebody is coming from a microbiology background or somebody’s coming from a mathematics background, so we want to show them the video more into there. But for that, we have to get some video-sto-generator algorithms to work recommendation system. That will be the first thing. The second thing, I think we have discussed this as well. I want to understand, like, which kind of videos people are more interested to watch, yeah, and what is, yes, yes, and what is the time duration? There is another concept now, like, because of this TikTok generation, I feel people are losing their concentration very, very fast. They’re focused like, you impress them into 20 seconds, yeah, or 40 seconds, and otherwise, they will just go to the other thing. Yeah, so we don’t have, I want kind of instant gratification on Stem Cognito, right? People have to sit for at least a few minutes and actually learn something, yes? Okay, so we can have an idea on all our videos over a period of time. What was the duration they kept on, like, and what was the duration they left the video? Yeah, and which subject it changed very on, which subject and what was the topic? So we can also understand the behavior, and this can also become a good mathematical, statistical, and science communication paper, yeah, where we can talk about how exactly this can happen. And the third thing is, we are still working on. Now we have, like, analytics tools, which we are using to understand how we can make our content better. Like, and we also, and find seeing from which countries and which continents people are coming to Stem Cognito, and which videos they are liking more, and what exactly they are expecting, or we can at which time they’re actually watching videos, because we’ve seen that many people do actually watch videos during their, apparently, working hours, right? Yes, super impressive. I think, uh, because people, like, when they are doing research, if you are watching something, and I feel people want to listen to something when they’re doing writing something or experiment. So now we are offering them a better alternative. Instead of watching some other entertainment, you can watch science communication videos, or you can also watch or listen to podcasts, and, you know, interviews. We have very diverse topics. Like, we started with robotics, microbiology, and we’ve gone to even dinosaurs. Yeah, like, we have discussed many different things in our interviews, and in the future also, we hope we will get all these diversities. There’s more to come, for sure. Yes, yeah. So I hope these things, like, we can still enhance whatever we have, and try to give them a platform. And another reason for me, as a mathematician or as a researcher, why I like this Stem Cognito is, I feel now, even if you are specific to a very minute field, like oil and gas, or I like some even mathematical modeling in oil and gas, you can find, now there are hundreds of papers coming every year. Like, we also did this study, like, we had two, three, some three-point-something million.

In 2020, even if you take a small field, you also have hundreds of papers, which is not possible for everybody to read and understand. For each paper, you need some time. But if you give the author the opportunity to explain whatever they did in six minutes or ten minutes’ time, the other researchers, like me, if someone else has done some mathematical modeling work, I am just listening to him firsthand. The author knows the paper best, I feel. He knows the best thing. So if he is explaining what exactly he has done, then I don’t have to spend more time. If I listen to ten videos, I can shortlist three which are more closely related to my work, and I can focus only on those papers. I have listened to them, I have seen their presentation, then I can go to their paper. And in the future, I feel it is the need. I think more journals will also now want graphical abstracts, and I feel very soon they will demand videos explaining your research, and not just writing. I mean, if you write, I mean, I’m a writer myself, if I write, I can take all the time that I want to think about the words I’m using. Well, if I talk, I talk in the moment. I can’t just go back and erase what I said because I already said it. So yeah, I think it’s a really important skill to actually talk about your research, explain it in one go, in an easy language to anyone. Yes, okay, so one last question I have about this Stem Cognito team for you is, I guess in mathematics, you’re surrounded by many male colleagues. I mean, I don’t know many females or women scientists that work in the mathematics field. So I assume that it’s very male-heavy in your field, in your department. But now in Stem Cognito, you’re actually working together with only females. How does that work for you? Is there a difference for you? Yes, I think I’m learning a lot with the Stem Cognito team. The first thing is the beauty of the Stem Cognito team is we are very diverse. We are from very different places. We don’t belong to one particular country. We are from different countries and different backgrounds. And when I’m getting experience of Germany, what exactly the people in Germany work things, and working with people from Malaysia and the UK, France, Australia, so I’m getting different perspectives. And how exactly the work is, like deadlines, if I talk about how exactly we plan things. And especially when you and Marta are taking leads in designing many things, it’s really a learning process for me. And it’s enhancing me as a researcher, how to communicate.

Yes, and we have a few best communicators with us, and also I am dealing with the interview process, part of this Stem Cognito interview. So I’m listening to all these different researchers while working on this project and this thing. It is helping me how to express myself, my research thoughts, in a simple language so that it can reach a broader audience and motivate people. I think we need more female leaders, and it will be very much fun to work. We have one of the best work environments. Although we are taking it as something else, we are just passionate about this project, but I feel even in the future we can enhance this and work, and this can be an example to anybody who thinks this is not possible. It is possible, and we are doing this exactly. Yes, and there’s so much more to come, that’s for sure. So good, yes, such a good album. Thank you. That has been really, really interesting and thoughtful. So at the end, we always have our last five random questions. Are you ready for them? Okay, I’ll mix them up a bit just to surprise you, because I don’t want to be too much telling you already what you should answer. Okay, so the first question might be easy for you: What was your favorite subject at school next to mathematics? My favorite subject was physics. Physics, okay. I feel these subjects come with good teachers. My physics teacher was very good. He made everybody in our class attentive to the subject, and when you understand the subject, I think that subject becomes a favorite. That’s my simple philosophy. Yeah, that’s true, that’s true. All right, and the next question is: What are you truly passionate about in one sentence? I’m truly passionate about cooking. Wow.

Amazing! So, we know where we’re going, and that’s Stem Cognito’s first anniversary, right in your kitchen. Yes, that’s the idea. I always want you to go to Malaysia. Amazing, yeah, definitely. And it’s my dream to someday open my own restaurant. When is it going to happen? I’m really passionate about cooking different things, not only Indian food. I want to try everything. So, that probably also answered the next question: What do you do in your free time? Cooking, cooking. Okay, that is so interesting. I did not know that. Amazing. Okay, and if you wrote a book, what would be the topic of your book? Mathematical modeling approach? No, I had another idea, like mathematical modeling in day-to-day life. I want to give people examples of where we are using it in small applications and how we are using it, how you can build. Okay, that’s really helpful, really thoughtful. Okay, and the last question is: Which skill do you teach your children on a daily basis? What do you think kids really need to know these days? I think now, I want them to be good at one particular aspect of anything. It can be cooking. I want them to have at least some kind of skill. My son is very passionate about painting, so I want him to, if he wants to make a career in painting, I’m okay. I feel even in the future, people will be more going into this kind of streams. In the last 20-30 years, there were more formal education-oriented. Everybody used to get a degree, then apply for jobs and these things. But now, I think people are exploring more varieties. After school, they start doing their projects. I just tell them that whatever they are doing, just take one skill. At least one person should have one skill on which they can survive. It can be painting, cooking, anything. It’s up to them. Yeah, I just, even writing, yeah. For me, I feel writing is one of the most crucial tasks to communicate. Yes, yes. I wouldn’t even just bring it down to writing. I think communication, in general, is crucial, especially these days, whether it’s texting, writing, or talking. Communication, in general, helps you with everything, with people’s relationships, in science, anywhere. Yes, and I was actually listening to this podcast the other day where one of my favorite science communicators, Emily Graslie, she said she always makes sure that or she thinks that an educated kid should learn how to survive on a low budget for cooking because then they can always feed themselves for little money, and that’s just such an essential skill. Yes, so yeah, I agree with what you said. Okay, everyone, this has been a real pleasure. It has been an absolute blast. Thank you so much for taking the time and explaining. Nothing, thank you for modeling to me. I finally understood some of it. I don’t want to think I’m fully involved yet, but I’m getting there. So thank you so much, and thank you all for listening. Talk to you soon. Thank you, bye.

Thank you for watching with Stem Cognito. Find more videos using the search box or the drop-down menus above. If you think there’s something wrong with this video, please use the report button to inform the Stem Cognito team. Questions about the video content should be directed to the researcher. You can find their details below. Go to our submission pages to find out how to submit your own video, and don’t forget to follow us on social media.