Environmental Sustainability with Process Mining, featuring Anton Ehret, Processand

Welcome. Welcome. Welcome. Welcome to the mining your business podcast show all about process mining data science and advanced business analytics. As you can tell, I'm not Patrick, but I'm Anton and today I will be discussing environmental sustainability. What I wrote in my thesis and how process mining can help you in reaching your sustainability goals. Let's get into it. Sustainability is a massive topic. After all, we only have one planet and we need to look after it. Some people are developing new greener energy sources, some people are gluing themselves to the roads. We are all, you know, after we are process mining engineers, so our approach towards a better and greener future will, you know, you guessed it, go through process mining. And to talk about how process mining can assist with improving our, you know, environmental sustainability. We have invited our colleague Anton Er. Anton, welcome to our show. It's, you know, it's very good to have you here and have you also talk as a environmental expert of process mining and it's coming from our company. So once again, welcome. Thank you very much. I'm super happy to be here and thank you for the invitation. We love having our own. So Anton, we usually start with a little introduction and therefore I don't want to skip this either. So tell us a bit about yourself and well, how did you end up in a mining or business podcast? So, yeah, my name is Anton. As you already said correctly, I live here in Munich and work a process and as a data scientist, apart from work, I really like to ride my bike and eat. Yeah, how did I end up with this topic? Just sustainability in general, whether it's social or environmental is a topic of high concern to me and with the urgency that comes becomes more and more apparent to do something about our environmental pollution and how it will affect us all, no matter if we are first word country or super rich or poor, it will affect us all at some point. And in our day to day life, when we do process mining or process optimization, we generally always measure a certain metric and then try to minimize that. So usually it's time or cost and with the problems that will come with increase in climate change, we will also have to minimize other things like waste. And especially environmental pollution and greenhouse gas emissions. And that's what I kind of started with this whole journey of combining sustainability and process mining. Now, was this a topic that you studied somewhere or is this just generally out of your own or maybe even self preservation? Like if nobody does anything or I can contribute a little bit, I'm going to do it just to save ourselves as a world. Well, I've done my master's degree in business intelligence and everything was always about, you know, I don't know, fraud detection. As I said, minimizing and optimizing some metric that is somehow connected to economic value. And thinking about, okay, we're, we have the pleasure of most of the times, these problems being problems of like big corporations that want to get richer and richer. And that's fine. That's cool. But I want to also do different things with my work and contribute to a better living, not only in terms of like we are all going to get richer and get nicer cars, but actually contribute to better world and just do something that is more than making people richer or reducing costs. That was actually my drive. And then I started my master thesis with digging deeper. And then now I'm working on how to get like from the scientific perspective into an actual application. And there are already really nice cases here. Now, wouldn't you say that if a corporation that's improving their processes is getting more efficient by product of this efficiency would be that they would be probably a bit more environmental friendly or something because ultimately if you are more efficient, you're spending less time or something that means less resources. And that by its own should sort of facilitate this degree of less pollution ultimately. Yes, of course, there are many things like, let's say, okay, we do order bundling or use cases like this, right? Help us in an economic case or productivity, which is also going to help us with, I don't know, throughput time and cost optimization. And then a byproduct might be that we have less pollution, we have less packaging and these kinds of things. So yes, definitely. But then there is also cases in which the environmental component and the economic component are actually contradicting each other. For example, let's improve shipping time. So we're going to fly everything, right? And so it can be that these cases also improve sustainability as a byproduct. But as we see with our current development, we're basically not going to meet any of our climate targets, neither governments nor companies. So we're going to have to go through more drastic measures and where we actually not have sustainability as a byproduct, but as our main improvement target. Now, what would you say are the main issues? Why process mining in general in this business process management? Isn't utilized as a go to platform or go to approach when optimizing the carbon footprint we are creating. So what are some of the setbacks or hurdles that we have in, you know, saying that every company in the world should just use any process mining solution. There is on the market to optimize your processes so that they are more, let's say, greener or, you know, efficient with the way that they are polluting, let's say. Yeah. So I think there are multiple issues. And two, I think are the most important. First of all, it's a lack of incentive. And the second would be a lack of quality and quantity in data. So with lack of incentive, I mean, right now it's still too expensive to save, let's say emissions or it is not cost efficient enough or the penalty is not high enough for pollution. So that's one of the factors. And then the second is just the quality of data. For example, if we're talking about an ERP system, then it's enterprise resource, right, and resources are most of the times not measuring emissions. And then there are certain formulas with which we can quantify emissions. But for these, we also need other data. And that data is currently not so well maintained, at least from my experience talking to like about five companies. So reading a lot so it's practical and theoretical knowledge that I have here is that these systems, so the source systems, which are the basis of our process mining initiative are just not maintaining the data we need currently. But do you mean if we look at the ERP, we're looking at some sort of maybe in purchasing or something, or if we're looking at the activity of change price, do you mean like that activity by itself has some sort of environmental cost that we could potentially track or how do you mean that? Yeah, I mean, so we have like a few standard use cases, and I'm saying standard, it's not really standard because it's not being used as a standard right now, but these are like the low hanging fruits. So, for example, when you're talking about purchasing, then we could say, okay, inbound shipping emissions, right? Or if we're talking about production, then it would be the energy consumption and in production and talking about, for example, okay, we need to do rework on this certain material for it to actually be where we are. And we could say, okay, each hour we spend extra on that material consumes energy, how much energy does it consume? Where do we get our energy from? What CO2 is related to that one hour of work? And if we have that information in our ERP system, then it would just be an extra metric like we would have, for example, time spent, we could also have energy spent, and then we could have an CO2 equivalent. Yeah, I can imagine that even quantifying these is insanely difficult because there's just so many variables that would be coming into the equation here to put a number just like you would with time, which is just such a universal measurement that it doesn't really, it's so easy to map into process mining. What you did though is, and you already mentioned it, that you did a master thesis on the topic where you were actually testing this null hypothesis, which was saying that process mining cannot be utilized to measure and improve the environmental sustainability of a business process. And you were sort of trying to reject this hypothesis and come with an alternative. Tell us about your thesis. What was the main goal that you were trying to achieve there? Yeah, so when writing a scientific paper like that, you always want to, so I always imagine you have this wall of knowledge and you want to either put a stone on top or put a stone next to an existing stone. So the first thing I was like wanting to know, okay, what does this wall look like, right? So what is out there concerning environmental sustainability, copper sustainability, process mining, of course, or data driven process optimization in general, and then are there any like areas where these things overlap. And of course we have like environmental sustainability reporting and these kind of things and people have in their non-financial annual reporting also sometimes they talk about environmental sustainability and some companies more than others. But what I could hardly find was really the utilization of process mining, whether it's with solonus or program or whatever and being used for tracking and optimizing the emissions in a process and there was one book which talked about, okay, one could do this, but I couldn't find any use cases. This was about one and a half years ago and a lot has happened and been happening since then. I mean, solonus is also really active in this regard, but yeah, back then I couldn't find any, any real scientific evidence of this happening. So that's why I've put up this null hypothesis because that was the evidence that I could find in the data in the data or in scientific papers. There is no use cases of this, so it cannot be used this way. And then of course it was my goal to contradict this. So when you say you couldn't find it in the data, does that mean you looked at actual data or was this mostly that you just theorized about if this were to be done, how would you do it? So when I just talked about data, then it was the data in terms of scientific literature. And then I needed to find a partner with whom I could actually conduct my thesis because I wanted to make it a case study in that I can actually test whether it's possible or not. So I first had to do the theoretical part in testing what is there in terms of science and then maybe combining things that have not yet been used for in this regard, like, for example, activity based costing, yes, we have it for costs, but we don't really have it for emissions. And then different approaches top down bottom up. Yeah, and how these could be combined in this new setting and then testing the hypotheses on the case study. So what was it that you actually did in your in your thesis then? I want to do two things basically. First of all, I wanted to improve the thing which is most easily improved and can be improved. And then also so can be improved in terms of, okay, I can do this with with the partner that I have for my thesis. And then it should also have an impact, right? And when you look at the like greenhouse gas emissions, which is actually what's basically driving climate change in terms of the more greenhouse in our atmosphere greenhouse gas in our atmosphere, the more the temperature on earth is going to increase. And this has a lot of following effects on every human being or every natural being. And the biggest or the biggest greenhouse gas or the greenhouse gas with the highest amount that we have, this is weird saying. Yeah, like the most prominent greenhouse gas is CO2, just in terms of quantity, not especially not exactly in terms of greenhouse effect, but in terms of quantity. And then you have that different sectors like, for example, energy and these different sectors, which are the source of the CO2 or the greenhouse gas in general. And we have energy and transport here, which are like major major contributors or major emitters. And so what I had and then with my partner, luckily I could get their order to cash data. And so this was then very much related to transport in terms of outbound shipping emissions. So then I had what I needed, right, I had a low hanging fruit in terms of, OK, this is transport emissions. It's CO2, right, we have combustion engines, we have jet flights, jet planes and so on and so forth. So yeah, that's how I then ended up with quantifying the emissions in outbound shipping. And in essence, what you did is you were looking at the data from your partner and we're examining how they are shipping from, let's say, plant A to plant B and we're trying to quantify the impact. Yeah, exactly. So the total generated CO2. Yeah, if you want to be very precise, it was from that plant to the customer's location. So the ship to location basically was the customer. And what you had were transports, which could consist of one or multiple deliveries and multiple deliveries could be, or when delivery could also be multiple transports. So you already see that there's a little bit of, it can be complications here, but in the end, I have taken the mode of transport depends on which to which granularity you can go. Ideally, you know, if it's a truck, how much load the truck can take and therefore thereby you also know how much it will, how much energy it will burn per ton kilometer. Then you take the distance and then you can take the weight that it has loaded and this way you can actually calculate the emissions per transport, which you can then track back to your delivery and even to your case in this case, the order. Is it then a case of how do I pick the right mode of transport to get there or is it just that how can I optimize the space more effectively within the mode of transport that I've picked. Can I maybe talk about volume or weight or what are we actually trying to optimize here. So there's many, many things you can do the first step is just quantifying it. So once we've quantified, then we can conduct different analysis on it or we can, we don't even have to start with the analysis, we can already start with the reporting. So we report our emissions, which is something that the market in general nowadays and with the younger generation coming up will want more and more investors are interested in your emissions data. So potential recruits, so our customers, a lot of people are just interested in your emissions and it can also be done for your non-financial reporting, it's just super high value to have this. The next thing would then be okay, we go into optimization and we go, so we look okay, do we have maybe certain certain case that are always related to super high emissions, why is that the case. Can we maybe, can we maybe pack something more densely or is it just a high volume good that always needs to be transported with like a super large truck. And even though it only weighs two tons, it has to be transported with a 24 ton truck. And so that's like the analysis, which is like centered around how it is right now. You can also go into the future right and do like kind of simulation or assist in decision making and of course, when you have these different input variables that are weight, mode of transport and distance, you can kind of see okay, which of these variables has the highest impact on the outcome. And it is not, you don't have to be like a rocket scientist to guess this, but distance has the highest impact on your emissions. And so then it becomes quite clear that selecting the optimizing your distance, so in this case, minimizing the distance. So then it is again a minimization problem, right? Minimizing distance that has traveled is very, very important or is the low hanging fruit for optimizing your emissions in this case. And so it's a problem of selecting what to ship where from which point and then the second would be because you can't really change the weight unless you optimize packaging or you have lighter materials. And then the next thing would be transport mode, transport mode, right optimization and then it becomes also quite clear. The cheapest is train in terms of emissions, then it's the truck with the highest load capacity, and then it's the plane. And then shipping is a little bit of an extra case, like shipping via sea freight, sea freight. Right, sea freight, an extra case. It's like super high distance and relatively low emissions compared to the distance. But the oil that they run on and the, yeah, it's quite pollutant, but still, of course, many, many, many, many, many, many times better than flying. But it also has the, it's quite slow and you oftentimes have large pre-dissessing legs and legs afterwards. So when I talk about legs, I mean like transport legs, so you need to transport, for example, from your plan to the harbor and from the harbor to the next harbor and then from the harbor to your customer. So then it becomes a little bit complex and, yeah. So ultimately, given, or let's say assuming that you have all the data, and I will talk about the sources for this data in a little bit, ultimately what you want to end up with is some sort of a process where you, in this case, you have a shipping process, sort of. And you have mapped as a, let's say, your main variables or your main dimensions would be like type of, type of transport, the emissions that this transport is creating and ultimately also distance. So if you are transporting goods from A to B, you would see all the stops on the way, you would see what is the distance and so on. And basically your process connection would be sort of not in timestamps, but rather in the pollution that is creating correct. So now this, this is a step where I still want to and hope we some day get to that we really, let's talk about like, okay, we are going to get process fear, right? So we're going to get it like the whole supply chain from supplier to end of life at the customer and recycling and disposal. So at some point, I hope we don't only have transport emissions, but we're going to have material emissions from supplier side, inbound shipping emissions, then energy consumption during production, then intercompany shipping emissions, waste inventory in distribution, things like, okay, packaging recycling of material that gets used during production and these kind of things. Then we have output shipping emissions, then usage emissions at the customer, right. And then we get into these scopes, one, two, three and so on and so forth. And then even at the end of life of the product and disposal emissions or recycling emissions or reuse. So these are all things which we can still connect and put into our whole process model. So yeah, it's got many, many, many, many layers and there's a super high potential. But as always, we need to data for it. So you would ultimately end up with, let's say, a life cycle of a material and everything that comes into this material in terms of production, in terms of transportation, the usage from the moment that it was, I don't know, instead of raw material, dig out of the ground. Until the moment it was decomposed or something and put a number on it and eventually in each of these steps, try to optimize it so that the impact is as low as possible. Exactly. That would be my dream. Yeah. Well, Anton has a dream. I love that. But Anton, you mentioned a couple of times already that one of the biggest issues that you had in your thesis and that there is generally on the market with this topic is the availability of data. And how did you, how, how did you tackle this issue in your thesis? So in my case, I need the input parameter for quantifying the transport emissions distance, weight of transport and means of transport and means of transport is also just another saying for I need the ton kilometer emissions. Per kilometer traveled and tons transported and just the easiest way to get this is by mode of transport because you know a certain mode of certain vehicle will burn this and that much fuel dependent on depending on how much load it is carrying. So these ton kilometers emissions per ton kilometer, there are standards like black or dean or even ISO standards. So if you have the mode of transport, you get to these ton kilometers. The next issue is then the distance. Now, this is theoretically there are fields in the transport tables, which would have this if they were maintained. A lot of times they're not actually maintained. So then you have to take ship from ship to or even go to plant and your vendor information. And for this, I needed to use an API. I myself use the Google distance matrix where I plugged in my data and it gave me the kilometers depending on the mode of transport. There are also alternatives, for example, climatic, which do this for you and then the last thing was the weight and this is something which is actually in my experience quite well maintained because you have your material mass data and even if it's not maintained in the delivery or transport, you can still get it from your material mass data by just multiplying it with the quantity. So yeah, this is for this specific case. Now, is it that you always know for take inbound deliveries, for example, and do you always know where exactly this piece came from and with what modes of transport, it was it was transported to you or do you just have to kind of roll the dice and guess if you don't know. So now now we get into the area where it gets a little bit spicy, right? You have vendor master data and SAP theoretically, you have also vendor master data and other ERP systems, but whether that's actually where it came from, you do not really know maybe it's under delivery, but you also don't know for sure. So hopefully you have it, might be you don't have it. Talking about location and then talking about means of transport, same thing. Now I'm looking for the word. It's like in German, it's up there and these things, you know what I mean Patrick. Basically, who's responsible for the shipment at what point the shipment becomes property of the buyer? And I'm just looking for the word right now, but doesn't matter. Yeah, I don't think there's a word in English for it. Looking in Patrick's face. I'm struggling. So this is also something like when does the shipment actually or when does the good actually become the ownership of the buyer? Let's say if it's my ownership as soon as it leaves the customer, then you have more control and you might have the state. If you become the owner as soon as it arrives at yours, at your place, then you have less control of it. So yeah, in terms, it's the word. So this also plays a role and if you don't have the data, you have to assume. And as I talked about before, the biggest impact is distance. And if you know, okay, I get it by road because usually that's something you usually know whether it's road or a f-ride or a train. And then if it's actually a 12 ton or 24 ton truck does not make as much of a difference. So you then you get like into these realms of 10% right? And I think having something with 10% variance is still better than having nothing. So yeah. So clearly the solution is to move your customers closer to your plant, right? Exactly. Okay, so actually looking at some of these assumptions, is that like a or in order to tackle these assumptions in order to get more data, right? Because I think at some point, you know, there are murmurs or rumors that, you know, at some point the EU is going to start regulating that this information be provided or that this information has to be collected and all this things. Where do you think is the biggest point that people or what companies should start tracking this data is the logistics inbound outbound? Like, where do you think is the biggest point of to start collecting this data? Are we talking about transport emissions or general emissions? I mean, if you're a company, right? And you're looking to actually quantify this. Who do you say you start tracking the CO2 emissions now because we need it? So let's say we are a buyer, so we're talking about inbound. And if we have the position to choose from different suppliers, then we can just tell them, hey, tell us your material emissions, give us the data. And we will consider you in our supplier selection with more, we would rather consider you than somebody else. The cynic in me tells me that there might be a little bit of generous submissions of their climate footprint, I think. Do you worry about that? It's going to be regulated more and more. People are text based on it. There's hopefully going to be more expensive climate certificates soon. People are going to have their reporting. So I hope there's always room for fraud, right? Same with financial reporting. People are always going to try and do it, but I just hope that it's going to be regulated to an extent where this is not the case. Yeah, and even if everybody fakes their numbers by a little bit, then the result in the end is to say, of course, I don't want this, and I don't think it's correct to do it. But I think this exchange of information has to happen, whether it's material emissions, or if it's just the actual emissions in production, or even if it's just, okay, just give us the correct starting point, give us the correct vehicle. And we do the rest. This exchange has to happen, but it's, of course, very hard. And this is also something where I think data-driven quantification helps a lot, because maybe you could push API or anything like that. So then there's also a low risk for the provider of the data, because they still have control, but we still get the data. Yeah, and there already are agencies today that are providing this information. It's generalized. It's maybe not for material, but it's already a good starting point to provide you with what ESG rating, do some of the vendors, some of the companies in the market have. And these are very simple metrics that you can incorporate into something that Anton actually mentioned, which is decision-ing with regards to, let's say, sustainability score in order to select the vendor that you would prefer to meet your own environment to go as a company. And as a matter of fact, I was even part of the project where we did incorporate this into what we called a vendor scorecard where one of the metrics that put a certain weight on the index on the final score of the vendor was also this ESG rating. So these things are already happening, and I also am very happy to see that. Next question I actually had, I still want to get a little bit back to your thesis and that would be, what was actually your outcome then? So you had some issues, but you had a clear goal which wanted to test, what did you end up with? So I was able to prove that my alternative hypothesis, which in this case was, okay, it is possible to use process mining for sustainability optimization of processes. I could prove that my alternative hypothesis actually is a true within the realm of my research. Because I have quantified the emissions and I have shown different means via descriptive analysis, prescriptive analysis, and so on and so forth how one could take this to optimize i.e. minimize the emissions in transport. But then of course it is, as I said, true within the realm of my research because I had to add certain data I had to take assumptions and it was also limited to the time frame of my research and the resources I had. So in the end, could you detect a significant or based on the choices that a business can make, can they reduce their CO2 footprint significantly or is it not we talking like 0.1% or something? No, it's possible to reduce the significantly. Do you have any numbers? No. But yeah, significant. I have actually another question. You mentioned that you worked on this thesis with a partner, which is a real company and you had a real real data. You also mentioned that at the end of this, you also presented to the board to the C level managers of the company. What was their reception of your results? Yep. They were amazed. I think generally management always likes to see numbers. If you tell them, hey, I can make it better for you as I just told Patrick is going to improve it significantly and they're going to be like, aha, but by how much and then you show them the number on the board, right? And you tell them, okay, this is how we do it. And we have this logic implemented. It works. And then you use this logic on all cases. And you have a number. And what then is the issue? Okay, nice. We're going to reduce our emissions. And then you have to still sadly put monetary value behind it. Because return on investment is still kind of hard to calculate if we have, let's say, accounts receivable accounts payable use cases, something like that. And we improve working capital. It's very straightforward. This initiative is going to cost us an investment of so and so much. I don't know, 100,000, but we're going to improve, let's say duplicate payments. And that this way, we're going to get a return on the west of 200 K. And okay, done. I'm going to buy it. But with limited resources that companies have, then you have to make an investment decision into something that's going to be giving me a quantifiable return on invest versus something that's going to maybe make my. And I'm talking in now in the mind of a manager, right? Not in my mind. Okay, this might improve our sustainability rating. Maybe we will get a few more recruits or few more investments. But I don't really know, whereas with these other use cases or standard use cases, you always have to put a quantifiable monetary value. Right. And it's always the return on invest on trying to reduce your emissions is maybe in like 100 years, we don't have famine and drought and floods and everyone dies. Right. So yes, it's or or if we get a move on, we're going to be text less. Right. We're going to like if if emissions are text heavier, it will have an higher impact. And then you can. Or if you if you have more expensive emissions certificate or less emissions certificates in the market, then you can actually behind every ton of CO2 emitted, you can put a monetary value. And this has to be high enough for it to have an impact on decisions by management and also high enough for that the cost of polluting is more than the cost of saving energy. Right. So this is what you were talking about at the beginning that the lack of incentive is is not there or that there's a lot of incentives. So basically economy 101, you have to say some incentive to actually do something. If you know, if the gasoline is costs one euro cent per liter, well, then you're just going to drive everywhere because why would you want to think about it? Very interesting. One thing I want to add, sorry Jacob, it's always easy for me, you know, to be here and rant about these things. But this is also something that we also need to work on. And I'm not just talking about us data scientists, but like in general, we also have to provide the technology to actually be able to reduce the emissions. And it's just in some books, it's being called the green premium, which is the extra that the other premium that you have to pay for the green alternative. And right now that premium is so high that it's most of the times on worth it. So it's again coming back to basic economy. Right. So is it who drives it? Is it actually the demand or the other supply that is driving it, but we need both me to hire demand for sustainability, but we also need more means of providing sustainable alternatives that are affordable. Now you actually took this initiative to a next step. And we talked about it in one of the previous episodes that there was something called a hackathon, which was organized by Salonis. And you participated in this with a company in Orbremza. Now I'm wondering what was this about and how did you. What what could you bring on the table given the experience and the expertise that you already build up through working on the thesis on the real data and how did you push it forward. So, yeah, this is actually a really cool initiative and a really cool movement that I see Salonis going towards that they push environmental sustainability. They do a lot of research on it. They are even probably going to provide some apps and the app store and so on and so forth. So this is really cool. I'm super happy about this. So this is one of the things I talked about before. Right. We need the technology to be actually able to drive these changes. And this is I think where Salonis is going in a super good direction. And one of the means of getting this technology that they develop to the customer are these hackathons and also one of the means to then develop these technologies further and to get knowledge from the customers. Now the second one was together with nine other companies and they said, okay, we have these four use cases that we know we have technology technological background for and that we can support you with and these were inbound outbound shipping and then material emissions. And, of course, the shipping emissions are a thing that I worked on a lot and have the experience. So that's where I came into play and together with Knoll Bremse. Yeah, started off the hackathon and work quite successful in our solution. So what was your solution? So and why was this successful? Why was it better than everyone else's? What we've done and this is I think where we had a real big advantage. A lot of other companies, their goal was to get to the point of sustainability measurement. So quantifying emissions, right? And this is already a big challenge. I mean, I've written a whole thesis about it, right? I've done over a few research on just this. And then also to have this measurement with a certain confidence that it's correct. And what we've done is we were very quick in our measurement of sustainability of our inbound shipping emissions. Talking about okay, the material that we buy from our suppliers. So we very quickly had the basis of our initiative and a lot of other participants, that was their main goal or that was what they reached in the end, right? And they've built a nice dashboard on it and you could see, okay, this supplier is maybe not has higher emissions associated to it. So a lot of participants, they got to the reporting, which is already great. You have at one day and a little bit over a week of preparation, maybe of talks and organizing your company to take part in the second one, but one day to get something like that, like that is huge. But we were quite quick with that and we also had a super good team that we then were able to go one step further and go into the, let's say value generation. So we now have this, this sustainability measurement in terms of we know what our transports cost in terms of CO2 emissions. And what we did then is we looked, what material do we need at what plant, right? And then we looked at, okay, we have, let's say, we have five suppliers which have in the past provided us with this material. And then we've broken the cost down in terms of monetary cost to the smallest unit that they send. And we've done the same to the emissions associated with that vendor. And then what we had, we had the cost per material for, like in monetary terms, but also in CO2 terms. And now this was really cool because now we knew, okay, we have, let's say, you're a person working in procurement and you have to decide, you have, you have the request for 100 tons of steel. Oh wait, that's maybe a little bit more, let's say one ton of steel. And you have to decide, where do I buy it from? No, I know my pricing conditions, I know I have two vendors which cost exactly the same. So now I would just say, okay, I will just get it from, I don't know, whoever, the one I like more when talking to him on the phone. But with what we have provided, we've actually next to the monetary value also given an emissions factor. And then you could see, okay, we sometimes have like very, very similar or even the same prices for the unit of material. But one has like three times as high emissions, of course, mostly coming down to the fact that they're further away. I don't know if that's something that the general purchaser considers and this makes a huge difference, right, because if you always get the money from that, get the material from that person, because maybe it's one cent cheaper or it's just as much as the other. Then this is just not in your consideration, but if you as soon as you include this emissions data, you can actually take it for your decision making. Then of course, you can retrospectively automate from that based on, right, you can say, okay, they cost the same. So I always take the one that's cheaper in terms of emissions. And then now I'm starting to dream again, the next thing that would be okay, we include, because we have the data from our supplier, we include material availability, we include the average shipping time because we know it from our throughput time in our process explorer. And we also include maybe what, what does, how quickly do we need it? And then we have four or five factors, which support our decision making, and then we can give these factors a wait. We can say, okay, we really want to consider highly our emissions, then pricing, then availability, and then throughput time or something like that. And then we can have a really, really nice data driven decision making, which can be suit to a certain extent, even automated. And we can even give that to our customer and tell the customer, okay, you can wait, and maybe you can wait one week longer, but it's going to be the same price, but half the emissions. And that's where I think, yeah, this is really cool. And I could even build up on the dream and say, and then you build a digital twin of this solution and run different scenarios on how efficient and how likely certain outcomes would be. And yeah, I'm all starting to fantasize as well right now. Exactly. And if you know you have this and these different, like we have these different metrics that impact our decision or maybe we give the decision decision down to our customer, but we can then also tell our supplier, hey, we are now looking at this. So maybe this is going to decide whether we purchase from you or from somebody else. And that has a knock on effect for everybody. Exactly. It's just, yeah, it's, it's threats, hopefully, and hopefully. So even with them with the data that we have right now, we can already start looking at that the climate impact and make decisions based on it that have some sort of positive impact on our CO2 bottom line. Yes. That sounds really, really fantastic. Anton, just hypothetically, where do you think all these efforts will go or where would you like personally to see it go in the near future? Talking about us as process and we have customers that request this from us, or I go to the customer and tell them about it and then the interest is coming. So I would like to implement this doesn't have to be transport emissions that can be many other use cases. And I would like to take everybody to take like initiative and go in the lead here to do this because the earlier do it, the better it's going to come anyways. And either you're going to have to pay the price of taxation, certificates, losing customers, losing investors, losing your health because if that world is so polluted and so warm that we're all going to have to wear masks and cannot grow crops anymore. That is going to be the final result. So very drastically speaking. So I would like love to take everyone to take an initiative and I think it's the sooner you start the better. If you start, if you don't have the data now start collecting the data so we can run analysis in a half a year. If you don't have the data now, we can always fill those gaps and by enriching the data with the most likely variables or assumptions and yeah. And then what else can I say? Thank you very much for joining us in the podcast and sharing the knowledge and the experience that you've built over the years in the topic of sustainability. Very close to my heart as well since I studied the smart building technologies and the life cycle cost, the environmental impact was always something that was only present. You just couldn't get around it because in building technologies you always know what type of material has what sort of impact because that's something that's measured. You have buildings that have certain certificates, how efficient they are and so on and it's really cool to see this mindset to be finding its way into process optimization as well. So I also think that this is a green field where we still have lots and lots to build to learn and to eventually improve. So thanks for sharing it and I hope that also it's resonated with you, your listeners and that you are already getting some ideas on how you could utilize this with your implementation with your initiative or just bring something new and new way on looking at your data, your processes because ultimately we are all sharing this planet so the more we can do the better. Anton, thank you very much for joining us. It was I think it was both our pleasure and hopefully we will bring you again soon to share in a few months or years what was the progress on the topic. Yeah, I'd love that. Thank you very much for having me. Thank you very much for listening and yeah, I'm also excited to see where this can go. For you, dear listeners, if you have any experience with implementing environmental metrics into process mining or in the processes in general, let us know on mining your business podcast at gmail.com or just reach out to us on LinkedIn. If you have any questions, do the same and otherwise we will be talking to you in two weeks of time with yet another episode of mining your business podcast. Thank you very much and yeah, talk to you soon. Bye bye. (upbeat music)