Skip to the main content.

Extended Producer Responsibility





Advanced Mechanical Recycling



Mixed Waste Sorting


Holistic Resource Systems





Deposit Return Systems


04 - A digital circular economy

"Digital solutions are a key enabler for the circular economy and will help to provide
transparency on how much of  what material quality is available where" 

Felix Flemming & Andile Dlamini




Read transcript below


Felix Flemming

senior vice president and Head of Digital at TOMRA Sorting
Trash Andile

Andile Dlamini

Head of Digital Solutions Recycling at TOMRA

Digitalization is transforming how many companies do business – from the ways they communicate, to the ways they create, produce, and distribute their products. The waste management industry is no different. In this week’s episode of TOMRA Talks Circular, we chat to Felix Flemming (senior vice president and Head of Digital at TOMRA Sorting) and Andile Dlamini (Head of Digital Solutions Recycling at TOMRA) about how big data is shaping the world of sorting and recycling. 

Show Notes

  • When we talk about digitalization in the circular economy, what do we mean? [01:48] 
  • If you don't turn into insights, you cannot take fact-based decisions, right? [07:15] 
  • So, I think the role that TOMRA played in this and we're still playing I believe is being the market leader and being the technology leader in this space. [14:26]
  • Cyber security and cyber threats are out there, and we read about them every day in the news. [16:33]
  • There is big culture change that's needed to actually understand the added value of that digitalization. [16:33]
  • Building an ecosystem of anyone who is involved in the recycling industry to make sure that information can flow and the circular economy can actually start to blossom. [19:55]
  • Overall, I believe we as an industry really need to accelerate this journey into the future. [21:16]
  • Final thoughts. [22:50] 


Mithu: Welcome to TOMRA Talk Circular where we explore how businesses, municipalities and governments are collaborating towards a circular economy. My name is Mithu Mohren. 

Lately it seems you can't leave the house without stumbling over a story or an opinion piece about the revolutionary technologies that are helping to transform how businesses communicate and interact with their customers. AI, for example, is everywhere, and the hype is justified, digitalization offers exciting possibilities across many different industries. One of those industries that you might not think of is waste management and specifically sorting and recycling. 

I recently read an article in the Harvard Business Review that talked about AI and the misconception that it only offers insights, and this makes sense to me. The article went on to say that AI gives us the user the opportunity to make much better decisions, so it's not only the insights, but actually using those insights to make decisions that we might otherwise not be able to make another way to look at it is that big data leads to truly or true business intelligent and who's not for that. 

Here to talk us through this extremely complex topic is Senior Vice President and head of digital at Tomra sorting Felix Fleming and Andile Dlamini, head of digital solutions recycling together, they are leading the digital journey that plays a central role in the transformation to a circular economy. Felix and Andile welcome to the program. 

Felix: Good morning, Mithu 

Andile: Thank you, Mithu. 

Mithu: So, let's start with the basics. When we talk about digitalization in the circular economy, what do we mean? What is in a tail? Felix, start with you. 

Felix: Yeah, maybe I start with this one and then Andile adds to it. So, when we talk about digitalization and data in the context of the circular economy, to me that is a key enabler to make the circular economy fly one day. it's not the solution to the circular economy and the challenges behind, but it is a requirement, a basic requirement and enabler to get this to go. So we need to connect information across the value chain. We need to understand what material is available, where and how. 

In TOMRA we introduced the concept of eight gaps around the circular economy that are relevant to close the gaps, and three out of those eight are actually directly related to digital solutions and data, big data insights like you mentioned from AI. That's the data gap. So generally, there is not enough information available across the whole value chain on what is where and then there is a quantity gap. 

So, do we have the right amount of material where if you operate a process in the industry and you need a constant feedstock, you need to understand what material is available where? And then the last gap that is relevant is the quality gap. 

What is the quality that comes out of the process at a certain point in the value chain and entering the next step? So those 3 gaps are basically addressed with digital solutions and data, big data AI as technologies and in that sense, it's an enabler for the circular economy to become alive and functioning and rotating and revolving if you think about it, which material is available where and in what quantities and quality. 

Andile: And to just zoom in into what you've just mentioned Felix, if for our listeners, if I think about the, the data gap, they may ask themselves, what do you mean by the data gap? So currently in our plans, looking at the value chain for themselves, our customers do not necessarily know what's coming into their plant and what they are pushing down to the downstream. So, there is that gap in the chain. There's also a gap within the plant itself, for instance, our sorters do not necessarily know what they are getting from the mechanical sorters in in the plants and if you think about the quantity gap where digitalization can help is increasing throughput in a plant, for instance, we can increase the machine availability and machine of time and thus increase throughput. 

So we are covering the quantity gap and finally the example for the quality gap. We can use digitalization to increase quality, we can use deep learning technology that could help us actually increase quality by sorting a bit more perfectly or bit more clean associated. 

Mithu: But that makes it much clearer. Thank you. So, we've talked a bit about how this work works exactly. How does it help the customer? 

Felix: I think it's if I start here on that question. If, uh, if you are a customer operating or a customer of Tomra, for example, operating a recycling facility or any other process actually. 

This whole process seems to you from the outside as if it is a black box, you have experienced people that operate it, you have a gut feeling. You have observations but overall, what is going on in the process every 2nd that it runs can be considered like a black box and that is actually the point where data and access to information provides you really a valuable feedback on what is actually going on in that moment at this time. 

Andile: Perhaps to just add to that and to give you a concrete example with one of our customers for recycling as part of Tomra Digital, we are producing these reports for the customer where we actually tell the customer how are they producing, what are they producing and what quality are they producing and do these reports together with the customer, we're able to see some abnormalities. 

In the way that they were producing and through analytics of those abnormalities, we could save the customer about €8000 worth of cost that would have would have happened if the customer had sent down downstream bad quality because what usually happens this bad quality comes back to the customer, and it needs to be resorted. So, all of that resource. This stage would have happened had we not had these digital solutions. I think this is a very clear example of how it truly we can help customers using digital solutions. 

Mithu: Ok and this kind of goes back to what I said at the beginning, that basically AI allows for a better decision and it's not just for lack of better wear, it's not just the insights that we're talking about, but these insights that lead to very good decisions that might not be possible otherwise Felix. 

Felix: So, let me add to that big data question that you had and AI pointing to these insights. Yes, in the end, as you can collect as much data as you want, if you don't generate insights from it and so, you turn data into information and that's what Andile highlighted with the example he gave. If you don't turn into insights, you cannot take fact-based decisions, right? Pure data for the sake of data is not worth a lot, but if you add intelligence to it, either artificial or actually human intelligence, you generate insights from it and that is the key in this. 

Andile: In addition to that, I read an article that said the waste industry as we know it right now, we know waste is a resource, right. But we are moving more and more towards direction of waste plus data or plus information is a resource. Let me make an example. So currently I can recycle, let's say a PT bottle our sorters will see a PT bottle. 

That's from a shampoo or that's from a Coca-Cola. That a bottle for instance. So, this is waste. And this is a resource. And this is good from the outside. You can be able to recycle this but with Ai. The technology that we're currently using at Tomra as well, we're now able to distinguish between PT bottles that are food and that is non-food, and this is like with information this is even better because we want to know separate these two. So, waste by itself.  We need to waste plus information or plus data to make it even bigger. 

Mithu: That makes sense I hadn't thought of it that way, and you might have actually answered my next question with the example that you just gave and that is how does digitalization actually lead or improve circularity, can you give some other examples? 

Felix: If we go back to what we said in the beginning, right, the circularity requires information on material that is available in the market. So, if you are a producer of new packaging material, for example, you're running a continuous process, you need a continuous flow of recycled material that comes into your plant basically to produce new packaging. 

Today, that information is not available at all, so you don't have a continuous information and a continuous supply of feedstock into your plant simply because there is no information on what is available where. So, in the end these solutions and the transparency on the information extracted from the data will help to make a circular economy actually flourish and start to rotate if you want so inner circle without that it's always. 

Let me try something here, let me try something there. We've proven this concept here, but to turn that into a continuous evolving and rotating material flow, you really need to have a continuous information on what is available where. It's, I think where the digital solutions will play a key role and if you then want to link that, so actually to individual objects or individual brands for example, and make sure that they are responsible for picking up their elements or paying for the recycling of their material, they put on the market, then the individual object information becomes in addition interesting and important, right, where did my bottle of Soda end up and was it recycled, is it back in the loop? 

On the bottles we might be easier with reverse vending machines of thermal collection. There you have an easier link to close that. But if you go towards a shampoo bottle, if you go towards a packaging of your bacon and you can think of all kind of packaging, then actually the recycling industry has a huge, yeah data gap. If you think about it. 

Andile: Yeah, that basically means we need to be able to track we need technologies. That are able to track this and meet you already mentioned the AI. There are others in the industry, for instance, such as such as the digital watermark which basically also allows for tracking. 

You can also use the fluorescence technology which also allows tracking everything points back to what Felix has just explained. How do I find out where the package is in the in the village? 

Mithu: So if I understand both of you correctly, this is a huge enabler of transparency. 

Felix: correct 

Mithu: OK, and that's where we want to get to in order to really make sure that we're getting into a closed loop as much as possible, because as you've said, Felix, if you don't know where the material is and we don't know where it's coming from, then it's difficult to put it back into the loop. 

Felix: I don't think it's just a one to, this is a have to. I mean, if we have to close the loop, we have to get to that level of transparency. 

Mithu: OK, I agree, so Andile, you just spoke about Watermark technology fluorescent markers, so this tracing technology, what are some of the other advancements that you're seeing in the digitalization space? 

Andile: One of the most exciting ones is the deep learning technology and like we already spoke about artificial intelligence and deep learning and basically how this helps. It provides transparency within the plant for instance, customers want to know what material is coming into their fund suppliers as supplying bills of plastics to recycling plants right now, it's just a matter of sampling. 

To know what is the quality of the materials that's being supplied to the plant, but with deep learning and using image recognition, the customer is able to know exactly what is coming in without on the end discipline. This is really important because our such as our machines, are working based on the material that is coming in. Also, it is very important for our suppliers to know what they're giving to the downstream, thirdly, as well within the plant. Although our sorters are everywhere in the plant and they are able to tell, we are able to tell our sorters what is running in the plant. 

There are spots where our sorters are blind, and this is where technology like deep learning through image recognition actually helps. 

Mithu: And this is also something that Tomra is working  

Andile: This is something that Tomra is working on together with the normal and absolute task for instance. 

Mithu: And Felix, you actually, you brought digitalization that's that was your task when you joined. I don't know how many years have you been with Tomra now? 

Felix: It's five years now, five years, almost like an old employee at Tomra as Tomra timescale. 

Mithu: Well, I don't know if I'd call you old. I think you're one of the freshest minds we have. What? What is Tomra’s role in all of this? 

So, I think the role that Tomra played in this and we're still playing I believe is being the market leader and being the technology leader in this space. We also almost had we were the ones that had to start this journey right and. We also had the, I would say, the power to start this journey and the leverage to do so. 

Actually, started to look at us and started to pick it up and started to move and we've seen over the last years also competition and other integrators and partners developing solutions and it just shows maybe we were early in the game, not in the big digitalization game that is going on for 1015 years. 

I think by now, but in the recycling industry, Tomra was definitely the, let's say, the ground breaker into that space and we'll keep pushing it and continue to do so. 

Mithu: And when you say connected, can you give me a concrete example what that means? 

Felix: So, in the past the sorters were. They're basically designed and built to sort, and they are still so today. But they generated information for the sorting task and once the sorting was done, that information was gone again. When I say connected, what I mean is these machines nowadays actually can communicate with their surroundings. They can communicate with our backends like Tomra insights in our case, and actually provide far more information. 

That they generate while sorting for other purposes to generate insights to create or to bring shed light into the black box that the processes as such or to combine it with other machines information and actually optimize the process. So when I talk about connectivity, this is what I mean. We need to pay attention. All that needs to happen in a secure way, right? Cybersecurity and cyber threats are out there, and we read about them every day in news, but that is something you need to consider when you build a solution like this and you start from the start with the security in mind and then you can start to connect equipment and use that connectivity to extract data, generate insights from that data and then use it for example fact based decision making or optimization or creating transparency in the circular economy context. 

Mithu: OK, and you've just mentioned one of the challenges are there other challenges that you see Andile that you see in this space? 

Andile: Yes, yes, there there's quite a lot of them matter, unfortunately. 

Mithu: As with anything new and developing, right so. 

Andile: Absolutely, absolutely. One of the biggest challenges we teach digitalization in general, especially the B2B industry, is the culture change so when we try to digitalize in an industry or within the company, we are always first of course concentrating on whatever the value proposition of the company is in terms of Tomra, we are selling our sorting machines and we are really good at that and they provide value to the customer and the minute you start at digitalizing. 

There is that big culture change that's needed to actually understand the added value of that digitalization and this is usually quite interesting within the company as well as with customers as well because digitalization or software solutions are not tangible, you can't touch them like you can touch the hardware which makes it like really difficult to also see value or also even pay for that value. 

So, one of the things that I actually, for me, one of the biggest obstacles where digitalizing in an industry is this culture change of really believing in what's the data, what the value that the data brings. This is the first challenge I see. 

The second challenge is that in in a customer planned in industry general, we spoke about the value chain, but also within the plant that say in a recycling plant there aren't only Tomra sorters, there are other machines that are in the plant and these machines and the Tomra we need to start talking to each other. If you want to bring the real value to customer and what usually happens of course Tomra like ourselves, we are starting with digitalization. 

We are started only with our sorters. So when Felix is talking about connected machines so far, mostly those are only Tomra sorters, even though we can connect to other machines that party machines. I think the keyword here is that we need the challenge is actually collaborating and we need to start collaborating and expanding our view, not only within Tomra but also with other sorters of third-party machines, Felix, you agreed 

Felix: I fully agree, actually, I would add to that in the end, this journey of a digital transformation, no one will manage to do by themselves on their own, so it will require collaboration, collaboration with their customers, collaboration with partners, even thinking about collaboration with potential competition around certain topics, developing standards on data exchange you can think very far in that sense. In the end, it's about building an ecosystem of anyone who is involved in the recycling industry to make sure that information can flow, and the circular economy can actually start to blossom. 

Mithu: OK, so sounds like a lot of as you've said Andilla, there are challenges, but it's all moving in the right direction and I believe we've just entered into a collaboration with the partners so that we can, we can make the solutions that we already have here Tomra even better and move much more quickly because as you've said Felix, we have to, we have to do what we can as quickly as possible to move towards a circular economy. 

Felix: It's with uh with anything that you start new, you start somewhere small and you start to build and you start to learn and we've announced actually publicly end of last year that we collaborate with a startup in the deep learning space that is particularly around material analysis. 

What Andile pointed out identifying objects in the material stream in the input in the output of a plant. It's poly perception and with them, we're trying to really accelerate that journey. 

So we're integrating them on the platform side, making sure their solutions is accessible also through our solutions and that the information flows seamlessly. It's the first step into that direction. I'm quite sure it's not going to be the last step into such a direction. Overall, I believe we as an industry really need to accelerate this journey into the future. 

Andile: Not only are we doing that Mithu collaborating with externals, we are also collaborating with our customers so. The way we develop our digital solutions is what we call this code developing with our customers. So, we have key customers that we actually develop solutions with. 

We deploy them on their on their sites. We get regular feedback and we come back and change according to that feedback before we put these solutions to them. So, this helps in going back to my first challenge of making digitalization or software solutions more tangible. 

So, if you develop something that if you could develop with the customer, I think it's this, this helps in bringing value and this is also what another form of collaboration. That we're doing. 

Mithu: OK. Thank you. So, I don't know if you had the opportunity to hear, we had a couple of podcasts ago. We had a young man. as a guest, and he echoed what many young people are concerned when I talk to my sons at the same thing, this sense of fear, of what's going on, the direction of the world and how everything is of the holding at the moment when you look at digitalization, which is what you both do, what could you say to those young people with digitalization and how that might help or rectify the situation? 

Felix: That is a tough one to answer, so maybe first of all, I actually agree. Maybe I wouldn't call it fear of the future. I'm concerned about the future. What is going on? And I think that's where, us as Tomra, playing a contribution in trying to solve that. I have kids myself, the all this one is a little younger than that young man that was on the on the podcast a few  series ago. 

But there is a concern and I think the digital solutions, I mean they have a in the in the 1st place, they actually help on the sustainability journey by saving energy by saving air consumption by saving. If you think of remote services saving CO2 by just not having to travel to the site, you can fix things remotely. 

So, you can generate more out of existing equipment by optimizing it further using the data that is at hand. So, there is a lot of things where digital solutions in the recycling industry in itself already help with reducing the footprint. Then at the same time, if we get a circular economy to fly, I think that is a major step forward in reducing our footprint and if any material we don't have to produce new, but we can reuse in, in our hierarchy that we talk about, right, reduce or was it reduced recycle? 

Mithu: Reduce, reuse and recycle, yeah. 

Felix: So, in our hierarchy, reduce, reuse and recycle anything. We can move up is a good thing and also, here I think that traceability will help right? 

If you know where a product is going or a packaging is going, you can actually change the course it takes, for example or you might be able to produce more local and close the circle in a more local space and think there is various ways where technology and digital solution can help into the future. 

They are for sure an enabler. They are probably not the solution in itself, but overall, I would say fear I think is wrong. It's having a concern and taking actions that's far more important than being afraid. 

Mithu: Yeah, I agree. 

Andile: Yeah, I would. I wouldn't also put it as fear actually put it as opportunity, especially for, the young people because they are the ones who are more used to these digital solutions. I mean, most of them their whole lives are online. I think when it comes to current challenges of the people adopting digital solutions in our industry is actually the people my age and older. So I think the young generation is really understands. That we need digitalization, make sure that's to make this better and they do know this is for their own good yes it is us, our generation and the past generations, actually messed this up for them, but I think they understand that we need these dishes or solution is to make the world better for that. 

Mithu: Thank you very much. You have certainly put my mind and heart at ease. So very enlightening conversation with both of you and I look forward to talking to you again. Thanks for coming on. 

Andile: Thank you very much, Mithu. 

Felix: Thank you, Mithu, It was a pleasure. 

Mithu: If you've enjoyed this podcast, please leave us a rating, subscribe and turn on notifications to comment on this episode. Visit 


Love these and want more?

Enter your email address below and we will let you know when we add new episodes.