Supercharged Manufacturing: Where Operational Excellence Meets Industrial Intelligence

Speaker A: Thanks for joining. Uh, this discussion emanates from, I think, all the things that we've seen in industry, all three of us and many other people that we have conversations with about how so many companies, and particularly in the manufacturing sector, not only in South Africa, but in Africa and the

Speaker B: rest of the world where they operate,

Speaker A: see the challenges that we face with, uh, performance and getting businesses to perform with best practices with technology, with artificial intelligence. That's becoming such a big topic nowadays. And I think companies are a little bit confused or many companies are confused. Where does this come from? How do they bring these different worlds together? And one of the purposes of this discussion is to try and unpack some of that, you know, unpack, um, why do companies struggle? Where does this come from and what are they going to do about it? Because you can't just implement technology and think you're going to see the results. And you can also not walk away from technology and think you're going to be competing with some of the companies that do have information available. So just to start off the conversation, um, I'll start with my background. Just everybody understands who's part of this conversation. Um, as I said, I'm Oren Brink. I'm a mechanical engineer, grew up in the pulp and paper industry and, um, worked in different challenging environments for a while. Uh, CCI, Growthcom became part of that about 25 years ago. We've been privileged to work all over the world in very complex manufacturing environments, many, many different industries. What we particularly are good at is operational excellence, supply chain excellence across the whole value chain. And that's where Hayden and I met each other. Um, but you were still in a management position in that industry. Maybe you want to run us through your background, if you don't mind.

Speaker B: Sure. So I'm Hiram. Um, I've got a background in electronic engineering, um, did a master's using neural networks. So that's where my interest in AI kind of started. Started, I guess. Um, and then a background, predominantly, a little bit in banking, but predominantly in supply chain and manufacturing, as Art says. Um, and that was with AB Init and um, SOB Miller. So I think where we met was. So I was a, uh, vertical operations director for Africa for Maltings, um, looking at five, uh, factories across Africa. And then we'd actually just built a new factory in Elroad. And we were going through commissioning and operational readiness. And uh, you can imagine that running the two together in parallel was am. Um, and that's where we kind of needed your help from an operational Excellence perspective and from operational readiness perspective. So that's.

Speaker A: And you are the, you are in the artificial intelligence world now.

Speaker B: Yeah.

Speaker A: So data AI.

Speaker B: Yeah. So when I left ab, I thought, you know, with, with the STEM background and, and having done, yeah math, uh, my studies, uh, using neural networks, I thought, let me get into that. It's pretty easy because we're all engineers so we understand how it works, um, and it comes quite easily. And, and now um, I'm getting to understand how to converge the two worlds. So manufacturing data and AI, um, and now operational excellence. Yeah.

Speaker A: Correct, Joan.

Speaker C: Yeah. So how did I end up with Data Profit? Basically? So I'm currently head of services at Data Profit where we focus basically on manufacturing and technology solutions, uh, for that industry. I started off as a mechanical engineering undergrad. Worked 11 years for big industrial company in South Africa. The end of my career I was tasked with the operations of making fertilizers and explosives. Um, and there we had a plan that's underperforming as you looted. How do we get places from underperformance to high performance? In that journey I discovered the most kind of critical factors being uh, people and systems. And what do I mean by systems? I got to use a lot of data, uh, to not just evaluate the performance of what the plant is doing, but also what is my systems delivering to me. And there a great passion for me was created through my careers understanding that if we can combine this crux of uh, people performance with data and system performance, you can really achieve high, uh, performing entities. And yeah, that led me through my career also then all the way to Data profit where ah, we actually now uh, at this pinnacle of solving from both improvement methodologies, taking people along for a ride, but also leveraging, leveraging the data, uh, and the data sources available on an operational entity to also unlock opportunities for performance. So yeah, that's why I sit here

Speaker B: today with you Jim. Thank you.

Speaker A: Because everybody has X amount of resources, but one company performs at a high level of efficiency compared um, to the next one. And when Hiram and I met each other, as he explained in industry, um, he was on a path where he had to get that particular business in maltings up to a certain level, uh, performance by engaging people, you know, which he carved out a path for. But what you guys just explained from a Data Profit point of view, as you're now sitting in a very different capacity looking at information through using very powerful technology and way down the line once you've got good analysis and I don't want to put words in your mouth but good analysis, then understand what that information tells you and then taking the next step towards something like artificial. So maybe what we can start with, we are both wanting the same thing. We are wanting to help businesses that struggle with performance to get the best out of their people, use the information at their disposal, uh, make better decisions and increase performance in the same way. And we want to do it in a sustainable fashion. So maybe what we can start with is your perspective would be great to tell us what do companies actually struggle with in this world of information right now? Can you maybe share your perspective on that?

Speaker B: No, I'll, I'll go first. Um, and then. Yeah. Ah, so Johan deals with them every day. So I think most of them that come to us. Right. Um, the first question is, um, I have outdated infrastructure. I have data that, you know, the quality of the data is not, not good. Um, and then I know there's an inherent, inherent, probably implied sunk cost if I'm going to get my data infrastructure right. Um, and then lastly, how do I integrate all of this so that I have one unified view of my data so that I can actually do something with that data? That's the first question that gets asked of us pretty much most of the time. So before you even go to AI, this is, this is the big thing. Um, and then secondly, you will have read probably only 15 to 20% of all AI applications kind of are successful. Um, and that's probably because a lot of manufacturers are in, um, kind of POC limbo. Um, but I think also the most important factor driving all of that is the lack of change management capability. Um, but understanding what change management in a manufacturing environment actually means and not say in a banking environment. So you're typically going to deal with things like SOPs, um, you know, daily routines and all of these manufacturing specific contextual things, um, which makes it very difficult to get ROI out of any technology that's deployed in manufacturing.

Speaker A: Can I just ask before we move on to you, Johan? So companies, information is not new. Um, AI is new and proper analytics is new. But companies have been, you know, when we normally start engaging a company about information, right down the shop, let's use a manufacturing company as an example. And this could be in a hospital, in a ward where the nursing staff are looking at certain numbers for them to make sure that they achieve certain things, you know, within a shift, all the way through to a bank where there's a production line producing certain processes. And what we are very familiar with is the manufacturing environment where Operators are producing a bottle of cool drink and they've got a machine running in front of a very complex machine, lots of information coming towards them. Where we would normally start that process is for people to understand what are they responsible for, what is information that they note down that the machine tells them. And then over the years companies became a little bit better in the sense that they started putting information out and putting on a dashboard. And you would see some factories that actually had or still have dashboards up, uh, in the factory environment where there's information available. Not all of them use it effectively, but it is there. Can you maybe just contextualize when you say your kind of information, proper data analytics, how does that compare to the old views of just having information on the dashboard?

Speaker B: Yeah, I mean the thing that comes to mind um, is probably one of the core practices that you often deal with is visualization. Um, another core practice would be root cause analysis or problem solving. Um, so typically when something goes wrong in a phase, factory recall analysis could take a long time. So it means people, different engineers and different people in different areas of the business collating or at least collecting a lot of data, analyzing that data and then trying to find the root cause. And typically by the time that happens, uh, a week has passed. Um, so, so just, just, just yeah,

Speaker A: it was this, this difference between when people in the past, maybe 10, 15 years ago started putting dashboards up and where we are today. Yes, I think it's a different place.

Speaker B: It's a different place. So, so you had spc for example, you can see your SPC charts um, in the control room, um, and, and then now you've got to deal with dashboards, um, and, and real time data. So it's a different world. Um, I think what's also slightly different um, is that if you look at the traditional uh, control theory that engineers use is they, they might be looking at two parameters and they looking at whether these parameters are controlled or not. Um, and then they are in control. At least the SPC charts and the charts in the control room indicate that they are only to find that quality or throughput at the end of the game isn't where they expected it to be. And that's by and large because um, these processes of manufacturing are not linear. They, they, they're non linear. And so when um, they change one or two of those parameters, they might change five or six parameters down the road. Um, and so what's changed is that our data is, enables them to see that effect at the end of the game. Yeah, um, which they couldn't see beforehand.

Speaker A: And just for the audience. People are not linear either. Um, um, you know we often employ a workforce to produce a certain product a certain time, but if somebody had a bad discussion with his friend the previous night, he might run the shift differently. Um, you know, so we'll probably almost never be, um, have complete linear processes where you put this in, you're going to get that out because there's variables, there's people involved, uh, there's information available. It's not a linear line to make this product and sell it at that profit because it always comes at a cost and at a cost of efficiency.

Speaker B: I think that on the positive side where things have really changed also is if you're doing a loss and waste analysis. So again you're going to get data from different sources and you're going to analyze that data. What AI, uh and data is doing is enabling you to do probably more accurate analysis, more precise analysis and then things like predictive analytics enabling you to do more sophisticated optimization. Um, so when you align it with the leans of the world, uh, you're going to see supercharged um, operational.

Speaker A: Johan, your view on what companies are struggling with in terms of information and the challenge thereof.

Speaker C: So if you quickly think about data, there's kind of two succinct sources. There's what a factory is telling you through census and there's this typically this cohort of data that's generated by governance processes. And what do I mean by governance process is typically quality control or shift reports. It's somewhere there's a business process enforcing somebody to take information down of a success of a process or the state of a process. And you can also immediately intuitively realize they're very different in their nature. The ones the sensor will measure whatever is in front of it, regardless of what's happening that day, good or bad. Where we have ah, kind of governance processes dictating it's very much many times also subjective in interpretation. So why am I saying all of this? Noting this is there's a lot of things where you almost have human out of the loop, where there's a lot of opportunity to see what that information can tell you. But a real richness only happens when you can collate that information with the process captured information downstream as well. And they typically now framing it in the challenge is where the challenge lies for many manufacturers because the sensors will measure. But getting the work practices and processes aligned to constantly measure your output of your process in a consistent fashion is a very big Challenge a simple fact of ah, we alluded to people are not linear just now. Your quality inspector today is not um, feeling so great, so he might miss a few quality defects in his visual uh, inspection after the shift. But there again data is your friend. If we can help create proxies or line of sight to inform what the quality of your data capturing is, it becomes a useful tool again. And I think what we see a lot of time in manufacturers is to get your data to a uh, state of health where you can actually start using it not just for advanced ML techniques, but just having trust in your data to make daily, weekly, monthly operational decisions as well. And I see that as the fundamental building block. Yes, we sold ML and AR techniques but a lot of our service led approach is helping you build a database of information and data you can actually trust and make decisions on. That's the first key challenge to solve for. Uh, running to AI from day one is the wrong place to go. Let's first check the veracity of the data set we have to work for and then have line of sight of improvement action. So now we leverage the continuous improvement mindset to improve your data sets that you collect as well. And the subsequent upside of that is just not to prepare you just for AI, but it helps all your internal processes, uh, uh, loss, loss analysis, your daily meetings, all the other processes that's dependent on data, um, as a consequence also then improves because of that. But the first kind of point of view is always let's get the data into a state where we can actually trust it.

Speaker A: Can we quickly talk about what kind of companies are struggling with this? You know, people can access, children can access information on their phones and it's analyzed to an extent. So if the world of work that you went, what kind of companies are struggling with information? And maybe we should talk about why. I know you've said some of the answer now. So what kind of companies are struggling with information? Maybe too much information or it's not integrated or it's not interpreted and why are they struggling? Can you give your perspective on that? Should we stop with you? Yeah.

Speaker C: So we have in our customer base from a maturity perspective you get operational mature, uh, and mature, but also data maturity. Everybody has struggles. It's just how ah, wide and how deep the struggles go. So what kind of companies are struggling with it? So some of our advanced customers is struggling with integration of data sets and systems. That's typically the struggle to get there because you might have a lot of information floating around. But how do you Consolidate it, overlay it or even combine it. That's one of the succinct challenges then on your lower level of maturity is just where do we actually spend either the capex or the effort to start collecting relevant data? And for you to start collecting relevant data, uh, you first need to know what's your goal or why are you even embarking on this journey to create, collect data or start creating information around this process. So I see that almost if I have to put two categories as the consolidation of data challenge, but then also where do you start collecting additional information if you don't have it on a system represented yet? And how do you create momentum behind it? Because at the end of the day it's a business, you need to understand your goals that you're chasing for the rest of the actions you take to also make sense downstream of that. Yeah.

Speaker B: So which is, which is why I think um, our partnership is key is um, depending on the level of maturity of manufacturing excellence for that particular client, they're often unsure of where to start the data journey. So one, just to make it tangible. One question I got was we're about to embark on our um, manufacturing class, manufacturing journey. Um, when do we implement the data journey? Do you implement them together? Do you implement the one first or the other one later? Typically the thinking is world class manufacturing, manufacturing excellence or Lean or whatever it is first and then other stuff later. And that's what we see a lot of them also struggling with. So in their journey of traditional operational excellence, where do they incorporate the data journey or the AI journey? Um, and that's not an easy question. So, so the different views out there that the data AI question could supercharge the traditional approach. There are views out there that if you don't do them together then you're not going to get that um, sustainable, um, reward, Um, I mean you look at maybe manufacturing excellence, so you've got sustainable practices and then you've got your KPI performance. So the whole idea is to get sustainable performance out of the plant, not just performance today and then tomorrow you don't know what happened. So there's that ah, thinking that if you incorporate or converge the two worlds, you are going to get sustainable performance out of your data. So that's a very tricky question. It's more of a strategic question.

Speaker A: Um, yeah, and I think, I think what a lot of people underestimate, um, a lot of companies is that change is difficult, changing people is difficult. So I'm the CEO, uh, and I'm confronted with improving the profit margins of the company, the market is there, I can sell my product, but we're not running very efficiently. So I, uh, think a lot of companies think that if we just spend X amount of money on tech and AI, it's going to solve those problems because it's difficult to change people. It's easier just to buy tech and AI in some people's perceptions and, and get some of the efficiencies through that. But not realizing that to make use of the technology and, and uh, something like artificial intelligence, you need people on board with the journey and you need the whole people, you need everybody to buy into. What are we going to do with this information? How does it empower us? People use the word operational excellence. A bit loose. Um, there are so many methodologies out there. Lean, Six sigma, um, total productive maintenance, total productive manufacturing reliability centered maintenance, Total quality management. There are all these methodologies that have been floating around for 50 years. Um, something like 5s, which is a very basic practice which I've been trying to implement in my home to big resistance of some of my family members. Um, is supposedly an easy thing to do. You know, if you want to have a neat, tidy, safe environment, make sure there's a place for everything and everything in its place. But companies struggle with it. Why? Because people are not linear. People don't just do what uh, you want them to do. And I guess that's where we get confronted so often with companies asking us, what do you do now? We do operational excellence. How do you kind of understand that? And when the conversation starts evolving, we have to go to the place where one of the real basics of the journey, which is that a lot of companies who are competitive understand that to get my profit margin up, uh, and is very often locked into being more efficient, so not necessarily cutting costs out of the business, like restructuring the business, retrenching, uh, a whole lot of people selling off printers and coffee machines. Not necessarily that, but actually take what you've got, the machines that you've got, the people that you've got got, and letting that run at much higher level of, higher levels of efficiency efficiencies. So what operational excellence, to us, you know, we'll talk about supply chain excellence just now means that, uh, the people who make the money in a manufacturing environment are the people on the shop floor, the ones touching the product, touching the machines, run the machines. And that happens seven days a week, 24 hours a day in most manufacturing environments. So operational excellence is not when the bosses tell people what to do. All the time. Operational excellence is when you empower the operator to work safely at 2 o' clock in the morning because he or she sees something unsafe and because they've bought into what this company does and what we are here for. And now my contribution towards the KPIs and the goals of the company, I am now respecting that situation and dealing with an unsafe situation or you know, somebody's not following the standard operating plan procedure or somebody is not doing the maintenance correctly. A, uh, boss or a team leader or a supervisor cannot be there all the time to manage those things. We have to empower the floor. So operational excellence to us means that the company's got goals and objectives and that is cascaded down, not forced down, cascaded down to the very lower level of the organization where you actually have an operator and an artisan and a quality person and a health and safety person are uh, working alongside each other and asking themselves, I'm not coming to work just for a salary, I'm here to fight for the business existence. So to do that is the process of what we call operational excellence. How do you get those people to engage properly, understand the numbers of the business, understand that at their level, which we call goal alignment and then making the goals visible so when they come to work in the morning or shift that they see the numbers, they ask the shift from before, where did you ah, operate? What did we lose? What are we going to make in the shift? And then they fight to make those targets. And you mentioned the word recourse analysis earlier on is when then something doesn't happen the way it's supposed to, is that team then will try and find out what happened here so that we can correct ourselves so we can still achieve the output. And so many companies misunderstand that when they talk recourse analysis, it is a tool that to hammer the people on the shop floor. Why did you make a mistake? That's not recourse analysis. Recourse analysis is empowering the people on the floor to do that themselves. And we're going to talk about the information and the impact of that just now. So when we talk about value chain excellence or supply chain excellence, you now look at the broader picture, the typical CEO picture. So not just the operations director picture or even the managing director picture, it's the whole CEO picture. So in other words, we trying to sell a product into the market. There's an inbound supply chain side of the business, there's the make part of the business, we manufacture the goods and services and then there's the outbound supply chain side. So what we often see is, and we want to talk about information when it comes to that as well, is let's say you fix the operational side of the business by implementing really good practices. Often the bottleneck for you not producing competitively is maybe in the inbound supply chain side. We're not planning properly, we're not procuring properly, um, we don't have a good sales and operations planning strategy. And then the bottleneck could be on the back end. You know, the logistics, the networking, distributing, um, customer engagement side of the business. So when we talk about value chain excellence or supply chain excellence, it is, um, what does good look like? If you look at a business, if it's a hospital, what is the very best hospital out there? And why are they so good? Is it the combination of people, leadership, information usage, technology? There's a bunch of things that make them successful. And so often companies, when they look at either the manufacturing side of the business or the supply chain side of the business, they don't change because they think they're good. Good in comparison with who. You need to benchmark yourself with the very best people out there. Then ask yourself, because today companies in Africa can compete with companies in South America and outperform it. And they do, because the same practices apply. Maybe cost of goods are cheaper here, uh, than over there, but the same efficiency rules apply. So why I'm giving this context is when we started talking about information, Hiram, you know, when we started engaging, uh, you know, some time ago, I was thinking, what has happened over the 10 years, last 10 years, since information has become more available? And certainly what we have seen in a company where the practices are quite mature, they absorb it, they love more information because they are now on top of their SPC charts, the statistical process control charge, short interval control, they're on top of it. You know, like if you look at a filler, a filling line in a brewery, um, the filler is the core of such a line. And the filler operator is a very key person in the that line. And that fellow operator has information coming at him or her that he or she needs to act on. So when information, if it's mature, the line's running very well, production is very stable. To give them more information helps them a lot. But in an, in an immature environment where everybody's firefighting now, you're putting all these, you buy electronic screens, you put them up and you just see people walking past them, not paying attention to them at all. If you Go and ask somebody what's the information on their board? They look at and say no, the foreman knows, um, or that guy knows because they disengage from it because all they're trying to do is survive. They're just trying to survive the shift. And they'll be. And the artisan is being pulled backwards and forwards all over the show because there's a breakdown there and a breakdown there. Don't give me more information. That's a lot of rubbish. But if you ask an artisan your job, God needs to have eight hours of work on it, he'll give it to you because that's what the foreman is expecting. I'm not saying everybody does that. There are some good companies that out there. But what, what, what we've certainly seen is that that hasn't changed a lot for us. If a company is immature and they struggling to add more information on there doesn't always help. But, but at the same time, and I want to talk about that further a little bit more for a little later is um, machines running well is dependent on preventive maintenance and preventive maintenance is dependent on us inspecting in the machine sometimes visually. And data uh, has made the inspections much more accurate. So you can plug sensors onto machines to give you more real time condition monitoring information which, which makes it easier. So even in a reactive environment, if you do that correctly it can empower the foreman to know what the condition of the machine is so we can plan when to bring that machine down and overall maintain. So certainly from our side, from operational excellence, uh, perspective. Perspective, we have seen the shift where companies sometimes use information as a cop out because they actually need to change the maturity of their practices. But they think it's easy to buy one, uh, hundred fifty million rands worth of technology, um, or actually go understand the maturity of the business and say if we are mature and our practices are running well, but the next step is to become better at information. Let's talk to a company like dataprofit so we can understand how to integrate our data, uh, use all the different sources, start understanding it better and then take the next step for them. It would be sensible. And I'm asking a question. If a company approaches you and say what is data profit all about? Uh, we need information. Artificial intelligence is our topic. How can you help us? How do you then deal with, if you look at that or and you say oh this is very immature, those immature practices, how do we go? So maybe if you can just give your perspective on um, operational excellence in the Older days. Now we've got information. Is there anything you want to add to what I've just said that that's make it make sense.

Speaker C: I'm very bull on this. I think what's hasn't changed and most probably will never change is enemy number one is variance. It's not and we can classify it a bit lower. It's variance. Sources of variance. If you don't know what the source is or the cause is, how do you even start to fight your way out of it basically. And then also what's very difficult for companies if you have so much variance coming in, that's firefighting. As you mentioned, you can have the most best intention solutions in place, but you might be finding noise, fighting noise instead of managing a trend. Very easy to say in retrospect while you're in the midst of this battle, it's not very easy to see. So step one with a company is let's also try and see from the bigger picture, CEO view, but in our space more operationally, what's your sources of noise and can the data tell the story to distinguish between just sources of noise or trends that's actually emerging and that also kind of builds to the Pareto view of the world. Can you actually see the 20% that's causing you 80% of the heartache? If we can make that connections for a customer and then instill some view that can help them manage to see if this noise is reducing? Because that becomes such a powerful tool. If you fix something, does it remain fixed? Because also let's be honest, many times a problem is solved only to creep back two months later and now we misdiagnose it as uh, something new because oh, but we did solve it, so it must be something different. Having that line of sight is an especially powerful tool. Also in a world where manufacturers is asked to be agile, think what that means. You cannot be agile if you have a lot of noise surrounding you because how do you even move your position in a space where you don't know what your levers are for control? Understanding noise, being able to classify which you can control or which you just have to live with is a very powerful mechanism for manufacturers to move forward.

Speaker A: Sure.

Speaker C: I think the journey for me always is based in reducing noise or reducing brands and using the most appropriate tool to your disposal. Sometimes it's work practices and data and information is not even a tool. Yeah, sometimes it's the combination of both and there I get excited. If you can combine people, process and technology, it becomes very ah, powerful Mechanism to not just identify noise or variance, but also systemically fix it and ensure it remains fixed and solved. Now you see that organization from a performance perspective, leaving its peers behind because, um, they've sustainably fixed stuff. But a side consequence of this in, yeah, I think it's a difficult one to articulate, but people that go through this journey gain so much process, knowledge and understanding of their own operations because you see it from a new perspective. And a new perspective is the data and information. Previously it's always just word of mouth. We struggled with this machine, we struggle with that machine. But if you actually start to see the trends, the data tells you, it becomes a very, very powerful tool to help yourself out of this situation.

Speaker A: So let's talk about that a little bit. Let's unpack that. My mentor who taught me maintenance, always used to say to me when we used to go into businesses to understand how mature are they with maintenance. Because we would typically get called by companies that are in trouble, um, they're not performing and they need first of all a uh, quick fix to get performance up to stay in business. But secondly is to do that in a sustainable way. So that our strategy as uh, uh, a management consulting company, the philosophy that we follow is we should extract at some point. We cannot be there forever otherwise we're not passing on the knowledge. So our strategy is to help collaborate on um, fixing that problem that exists. Why are we not performing, why the machines are running? It might be poor reliability, it might be poor production management, it might be a poor supply chain. But then put practices in that will help those teams to sustain that performance going forward. So uh, what we want to leave them with is competency and capability to learn the whole time. And then obviously combining information is to bring information in so that they are, uh, it's a more powerful way for them of seeing this. My mentor M taught me data does not lie. And especially with the accuracy of data today, you know, you can't be reliant on opinions anymore. The data does not lie. So let's talk about combining these tools. I think some companies see combining best practices and this very informative world that we now have through technology and availability of information as nice. But I think the words that you guys often use is actually critical. Can you give your perspectives on that? So, and maybe we should start with you. Why is it critical and not just a nice to have?

Speaker B: So there's, there's a thing we often talk about data or AI readiness. And I think for me there probably three key things for Perfect customer to be ready. It would be that they have a continuous, uh, improvement mindset. Um, so, and they have standardized processes and it talks to your point around if that's the thinking and that's their culture and their mindset, then the acceptance of data to use it for continuous improvement is quite easy. Um, secondly, they need interconnected systems. They need a good OT infrastructure, IT infrastructure to be able to use that data. Uh, and we've spoken about the infrastructure and the data and the normalized data and the unified view. That's, that's the technical side and then the uh, the other side that they need is that change management capability. So implementing change, as I said, in a manufacturing environment is difficult. They need to realize that when we bring in data we can do stuff that maybe the traditional Lean approach couldn't do. So, um, for instance, you might be able to enforce an sop, but we can check whether an SOP is actually being implemented or not. We can enforce measurable control plans, um, using data. So those are important. Um, and then why is it critical? It's because I think if you think again around loss and waste analysis, so you'd go into a business, you want to find the gaps, you do a lot of loss and waste analysis, data is going to give you a more accurate, more precise view more quickly where your losses are, where your losses are. Um, you're going to get a more sophisticated way around improvement. Um, and then, and then I think also to Johan's point, data enables you to remove the variability in the plant, both external and internal. So imagine you've got different operators coming across different ships doing or, or having their own interpretation of the plan, control plan. Can you imagine the variability in quality and throughput in the plant? Um, so it's critical to remove that and data can do that. Um, I think the other thing is, um, and Johannes kind of mentioned it around line of sight. So you get line of sight of your intern process that enables you to be more flexible, more agile. And then I think also you want to, I mean you want to boost, boost the, the traditional manufacturing excellence principles like pdca. You know, having the data enables you to, to plan or have a view of 20 or 30 of the key parameters that drive that process. Um, the data and the technology helps you to do the, do the check and the act quite quickly. Um, and then finally, um, you know, one of the tenets of Lean is empowering people. And, and the reason I say you need to converge the two is because imagine in a lean environment or a world class Manufacturing environment where you empower the people by automating certain processes and letting them use their brain to do more effective type work. Um, and in that way you're actually empowering and, and getting more engagement out of your workforce. From the bottom all the way through to the top. Um, and then finally I think in terms of alignment, I mean, we've spoken to Dino and we've often asked the question as to what made the likes of Sab mala really good. Um, was that thing you mentioned around goal alignment, um, and the cascading of KPIs, that's incredibly important when you implement technology or even data and AI. So, um, if we give you an example, uh, say you don't have that. And so you go into a customer and the customer's asking for a very specific use case like the eradication, scrap or defect. And then, um, when you're doing your change management, you ask them, so how significant is the KPI, um, compliance to your control plans or scrap, so your scrap percentage? Um, well, we're driven by volume. That's not such an important KPI. Well then how do you expect the technology to work? And if that's the case, often the question in our minds is where do you start? Do you change the operator's KPI or do you go back to the top and do the whole goal alignment exercise from scratch?

Speaker C: That's.

Speaker B: So it is a question that I was keeping for you, by the way, um, is that if you're going through an implementation of that, please start right at the beginning again or. And so that's why I think it's absolutely critical to combine these two worlds. Um, um, for me, that last point was, was the most important.

Speaker A: So just before, Johan, before we move to you, I think what, what you find out that what's quite important is that, um, we, when we walk, um, through a factory the first time, um, I mean, depending on the size of the factory, it could take us a couple of days to understand that's really going on. If somebody says to us, listen, we are not performing or performance is not sustainable, we've got quality issues or, you know, we've got a very active culture, whatever the problem is, um, before we put any solution up, because we don't have a blue and a red bull like in the Matrix that you just

Speaker B: swallow and your problems go away.

Speaker A: Um, people are not linear, so you can't swallow bull. Um, we, we always. And we don't believe in silver bullet solution. So we, we have to find out like you guys have to Find out what's going on, where's this business at, so we can identify what is the solution that they need to implement and let's collaborate on building that joint solution. When we do that the first time, we always need to understand what's the culture and what is the maturity of this organization. And so often you walk into a room where, like in a manufacturing environment, the numbers are posted on the walls. The, the, like the last 24 hours or the last couple of weeks and some of the charts are outdated. And um, but when you speak to the guys on the floor about the performance, I said, no, we know what we doing, but they're doing it informally. So the board is just up to keep the boss happy because the last couple of hours of the last two days have not been updated on the board. So immediately you get a feeling that there's a disconnect with the information. So to come back to your question about, do you start at the top? Do you start at the bottom? Bottom. If the people who make the money for this business, who touch the product and control the machines, don't have clarity on, um, what they're doing, how they're supposed to do it, how to do it in a standard way, so there's a little variance as possible. You know, it doesn't matter how we're going to do this KPI, uh, if that's not right, then we might not be making money because we're not controlling

Speaker B: what we're supposed to do.

Speaker A: So the question then is, how do we fix that? Do we start the con? Can we do that by starting conversations here or do we go back to the top? But that's, that's a whole lot of a big discussion. But if you've got people at the bottom of the organization not knowing what they are doing and exactly why and how, and doing it consistently in the same way across all the shifts, we're not efficient.

Speaker B: Yeah. And if, if, if you have the MD or manufacturing director saying, okay, I'm not sure I want to change the KPIs at the bottom to reflect scrap as an example. I'm just using that as an example, then, then you kind of got to go back to the whole start around the data and AI journey because it needs to be a strategic decision.

Speaker A: Sure.

Speaker B: What are your strategic objectives?

Speaker A: Yes.

Speaker B: Is AI, uh, even, or data even the way, uh, or the method of, of addressing that? And, and what are the key objectives you want to achieve as a business? Is it cost reduction? Uh, is it the improvement in capacity? Because you don't Want to build it another factory. So if that discussion hasn't been had, then you can have that lack of clarity at the bottom in terms of what categories uh, they should be chasing. And, and you'll quickly just, you will encounter it maybe halfway through the implementation or even through running the business. Because the guys are going to go, well, we don't have scrap KPIs. Why should we?

Speaker A: Yeah.

Speaker B: And the MD is going well, we actually don't need it. Then why did you implement the data?

Speaker A: Yeah, and I uh, want to ask you on a question, but before I do, sorry. Um, um, maybe one of the questions one should ask is if you've done the strategy for why do we need artificial intelligence? Or do we, or do we need more information? If the answer is yes, is the next question. Are there things that need to be put in place before we do that? So for instance, if I want to put technology in, um, am I ready? Am I ready for it? How do I get my people involved? What change management needs to be done? So probably part of the strategy, and I'm asking a question and giving the answer in the same sentence, probably part of the strategy should be as once the leadership team accepts and understands there is an opportunity in information as to then agree is what's the process to get there? Because if we are ready, we can do this, this and that. But if we're not ready, we first have to do do this and get that in place and get understanding and we'll talk about the solutions. Johan, can I ask you, um, how can we like I've heard you guys use the word supercharge. Supercharge and engine. Can you talk about that? Like how do we use, how do we use good information management to supercharge this manufacturing engine?

Speaker C: So I uh, always like to think in simple terms their metaphors a great

Speaker B: method to do that.

Speaker C: So I typically refer to the work practices at a company as the engine. Because if you think about it, if you have unhealthy work practices, this is the engine leaking oil. So if the metric is making power for that engine, the likelihood of you making consistent power is not very high. And that's a people process. Your people in the process, how mature are they? How good is this engine? Is it a nice V8 purring or is it uh, uh, 1200 or 1.2 barely started. If that's not in place, adding technology, and I call technology the supercharger or the turbocharger. Adding technology to a bad performing engine will only break that engine. In all likelihood, if you have good Practices. That's just what Hiram mentioned. If you have standardization and a continuous improvement mindset it that's a healthy engine. Now if you add technology now it gets exciting because now you add a turbocharger to an engine to make expensive exponentially more power out of the same fundamental base you have going at the place.

Speaker A: Is that the difference between noise and good analysis 100%.

Speaker C: Think of an engine that's not well maintained is a very noisy engine. Is not the kind of metaphor combines that this engine rattling away you don't know where to start looking. Adding a turbocharger to it will all likelihood increase the noise.

Speaker A: Yes.

Speaker C: And that's the danger of employing technologies you don't understand what you're fixing. With it, you're all likely going to create more variance sources of noise in the implementation. And it comes back to goal setting of clarity of business KPIs that you're chasing. Is my engine suited for this goal? Where it really gets exciting, Arendt is if you have the maturity as an organization to start designing your engine to accommodate technology a supercharger. Think in a mechanical world. If you design the engine to accommodate a supercharger, it's much better than retrofitting a turbocharger. It can still be a good solution. But you have that mentality of starting to design. Think about lean. Lean falls flat many times. Not because lean is not a great methodology. It's just sometimes the effort involved in doing it makes it unsustainable. Now let's design the lean implementation with where can we use technology to reduce the effort of the implementation cohorts? Now you start to see an engine that's being designed for supercharger and those opportunities and customers that gets that is to be honest, uh, the customers that will get the most most out of this combined journey. And that gets me very excited personally.

Speaker A: Can I quickly ask on that Is will there ever be a case where you will do one or the other? Because some companies will say, uh, you know, I don't need data and uh, I don't need information. You know, I need to go into the best practice, you know, lean, operational distance, whatever the case may be. Other companies might be of the opinion that, you know, stuff the best practices, you know, we just want information, information. Will you ever be in a situation where it would be one or the other? Or is it a question of how do you understand what you need from both, uh, worlds? What, what in your view is the solution?

Speaker C: I'll take the first element me. The one cannot live without the other. Is My honest opinion, it's always just the magnitude of how deep you need to go in each dimension given your current context and problem you face. But for me the people process side I your methodologies to improve is critical. You can't do away with it if you want to keep up in a competitive space. In my opinion today you cannot ignore the value add that can be done by including data information sources in your journey towards operational excellence or high performance. So yes you can, you can always do what you want. But you'll be missing out on all opportunity. Uh, so it needs to be both in my opinion. I don't see it will not be a strategically wise decision to be either or. But it's also in fairness due to reflection to see where the two worlds actually supercharges the initiative. Because if you do it in an unplanned fashion, we have opportunity to create more resistance and noise invariance. If you don't do it in a strategic minded fashion.

Speaker A: Did you have a view on that?

Speaker C: Yeah.

Speaker B: I'll talk about the specific risks of not combining these two worlds. Um, not now but m a little bit later. But, but essentially if you don't combine them I think it talks about sustainable improvement. So, so your world is about getting sustainable improvement. Not being able to improve your KPIs today and then tomorrow you don't know

Speaker A: how you got there.

Speaker B: Um and I think if that is the premise of what you do then you want sustainable improvement using the data as well. So you've got to embed the two. And then, and then I think also most importantly is the thing around people. So you know, get getting, getting uh, I, uh 4.0 implemented um, in an industrial environment. Environment is not easy. Um, and world class manufacturing or manufacturing excellence is actually premised on change. Like as a plant manager, if you don't know change management then how do you continuously improve? Because continuous improvement implies change. Uh, and, and how do you install those new standards and new practices every day or new ways of doing things? That's change management. So unless you combine, combine the two worlds, those are the two areas that I think you're going to fall on. And that's why you read that a lot of these AI and data implementations aren't returning the ROI that you need to return. So, so, so for example, ah, let's take root cause analysis. Root cause analysis. You might use data instead of world class manufacturing and the data will bring out some patterns or whatever it is around the root cause. But, but you might not. You, you might see the symptoms through the data. But you don't get down to the real root cause or um, if you implement only the data, you might miss the human impact on a particular production process. So you've got to do them together.

Speaker A: Yeah, this is a good conversation. I really hope the listeners get the value out of this.

Speaker C: But I do want to mention something that we've been saying it in very different ways. What I want to explicitly put it out again is this concept of meaningful work.

Speaker A: Orange.

Speaker C: Uh, you said it. How do we engage the people on the bottom of the floor? You said again, we need people to change their behavior. And I'll tell a simple story. If you start your operator, you make a part and you work very hard today, if you don't even know how much parts you made that day doesn't feel like a meaningful day, does it? But having the simple fact of having, uh, a counter, the simplest kind of metric to say, oh, I made a thousand parts today, yesterday I only made 800, I did better today than yesterday. All of a sudden you unlock this meaningful work mentality and I think that's the opportunity for you to make change sustainable. You want to make it meaningful. You make it meaningful by being able to track improvement.

Speaker A: Sure.

Speaker C: Or if you're going around wrong way, that makes it meaningful work. And I think that's kind of this golden thread tying all of this together. By visualizing the right stuff at the right time for the right people, you make your work on a daily basis meaningful. And that's the opportunity in front of us, uh, to do as well.

Speaker B: So that's a great point. And that's where I had a question for you. So, so globally from a data and AI perspective, one of the issues that manufacturers are having is when you saw the great resignation, um, after Covid, but the shortage of staff in Europe or workforce in Europe, skilled workforce and in the U.S. um, and then one of the tenets of what you guys do is an empowered workforce, an engaged workforce. In your view, from your experience around manufacturing excellence, can data and AI create a more engaged workforce? Um, can it in some way alleviate the problem around, uh, skill shortage? How do you, what's your view around

Speaker A: that whole issue that manufacturers I think always. Um, but I, but I think this, this whole picture that we are trying to sketch is not a one element picture, it's these different successes factors to it, the integrated picture. So absolutely, if information is made available to the right people at the right time for better decision making, it's going to make them more efficient. Um, will you be able to retain those people better probably if um, you know, if the companies, Because I think aam, you know, like you mentioned SAP before, how good they've been over the years with the competence acquisition process, building capability with the people and if people be, if the lines become, the manufacturing lines become so efficient that we actually have uh, too many people running too many operating line, we don't retrench those people, we redeploy them. Uh, I know SAP has done that often or companies like SAP has often done that. So redeploy those people into different roles to drive improvement even further forward. So I think the availability of information, information and having a more powerful um, view on information to the business can certainly make a difference. But it all comes back to everything we've discussed. Can you use it? Because if you make all this great information available, but like what Johannes just said, it's not meaningful or it's not at the wrong time or we use it for the wrong things, it's going to have the opposite effect. Or the engine becoming more noisy, it's going to have the opposite effect. So I think it's the bit, there's the dependencies in my view, you know, dependencies on good practice, there's dependencies of where you're on your maturity journey. So when you introduce information or AI, when do you do it? How do you do it? What is the change process you have to go through to actually make it work effectively?

Speaker B: And what's, I mean finally just linked to that. What, what's your view? Because I mean you read globally one of the impediments of one of the obstacles to AI, ah, is skills. Um, so when you think about skills from a data and AI perspective, what is your view around getting ready for that from operator level all the way to plant manager level. Um, so take the plant manager for example. Do they need to be ready and from what point of view? They don't need to be ready. AI, uh, pros I guess. But how does that differ from plant manager to operator slash engineer, maintenance, individual.

Speaker A: I'm going to take a wild guess and say that I, I think more information will almost always help but in different ways. Like at the shop floor, if you look at gender, Mark Yane's team and what they do with VR and, and, and um, training operators or helping them to put a gearbox together and using information to have the SOP executed in a more consistent fashion. That's more information at the right time, um, to the right person. It's almost like you're giving them a skill that they didn't have before, you know, so for me that makes sense. You know, in the right environment, giving them that kind of information at plant manager level, I think it will look different. You know, what is, um, how do they, are they proactive or reactive? If they are reactive, I think it will be dangerous to give them more information and because it's not necessarily going to help them. What they really need to do is run the daily management system better. Um, but if they're already proactive and you want to take, you want to skew up that plant manager to a different level, um, giving them the right information, more available so he doesn't have to look for it, doesn't have to dig around for it. It's more meaningful information so you can make better decisions I think will be meaningful for, but that plant manager needs to be ready for it, needs to be mature enough to handle it. And, and I think always, uh, I don't think we can ever get away from the, the basic skills that you're going to need to make this journey come alive like um, good people skills. You know, because we're still going to have to lead our teams and we still have to meet with our people in the morning and look them in the eye before we look at the dashboard. And if you've got a guy with an autocratic man management style, meeting with a team with great information might not have the same effect. So I don't think there's, there's certain things I think we will never get away with, move away from.

Speaker C: I just want to throw a small side note to it. AI is a destination, it doesn't need to be the destination. I mean there's so much value unlocking opportunities, just starting to use data, uh, information. We always kind of, uh, hold the pinnacle of the technology at the time and front of us. But uh, I just want to highlight there's so much information. Even if the end destination doesn't end up being a fancy AI solution doesn't mean we need to not embark in seeing what opportunities is in front of us. It's just a destination, it's not the destination.

Speaker A: At the end of the day, um, hopefully people listening to this will see the value in understanding the convergence of best practices through people. And how do you bring information into that world and how fast can you move? Can we maybe talk about where do you start? So let's say somebody's listening to this or looking at the saying, I see the value in information. There's so much information. I'm wanting, I can't see it. And they don't know where to start. Some people have to view, you know, maybe I need to spend a lot of money on analytics up front. Um, and what will this give you? Some people have the view that you need to plan properly and understand the stakeholders and business objectives and then you align. Can we maybe think about a client? Well, like we just mentioned it five minutes ago, let's talk about how we do that at the moment. Um, we would. When we engage a client who's got this need, they don't, they want information, they're not sure where to start. Can we talk about how do we start this process? You want to give your perspective, perspectives, and I'll add on to that.

Speaker B: Yeah, um, I mean I think, I think I mentioned it early on in the discussion that the uh, probably the perfect customer, uh, would um, be something. You know, we often talk about data and AR readiness. So it's a customer that has a continuous improvement mindset, standardized practices, or should I say standardized processes, and they continuously want to improve their processes. Johan always talks about people process and um, then technology, and he's spot on. So, so if a customer doesn't have that mindset, you've got to ask yourself, how ready are they for data and AI? So that's kind of the starting point. Um, and yeah, so, so we, we do the discovery which will do some kind of assessment and we'll kind of look, look at that. But I'll get to that in a moment. Um, data obviously, uh, is very important. So when we do a discovery, we're looking at the data extraction strategies, the data infrastructure, how disparities are their data sources, um, how, how uh, outdated are the data sources? Do they have hundreds of different PLCs? How difficult or how easy basically is it going to be to integrate all of this and create, create a good OT environment and a good IT environment where you can use that data. Um, and then ah, thirdly is the change management capability. Um, so those are probably the three areas that you need to consider, but in terms of where you're starting. So, um, we'll talk about the customer that we spoke about a couple of minutes ago. Um, in the context of our partnership, if got to go in and say, how's the technology going to deliver roi? How's the technology going to fix something in the business that isn't working properly and a nice way to start. I guess there's different ways to answer your question, you could start in many different ways, but I think from a Business case perspective would be to identify what the gaps are. Typically what you guys do would do a loss and waste analysis. You look at the data and try and identify the gaps, right? Um, and then you do our uh, assessment um, from a data perspective and you identify the gaps there, gaps in the infrastructure, gaps in the quality of the data. And then you look at that full picture and you say what gaps need to be closed. But often from your assessment you'll be able to identify that fixes both on your part and our part could increase or reduce cost costs by say, or increase capacity, say about 25, then you have a good solid business case. And for me that's the place to really start. Um, why is that a combined world? Um, because for example um, and I'm talking here about our uh, assessment which includes both the maturity from a world class manufacturing or manufacturing excellence perspective but also from a data perspective, uh, infrastructure perspective is that let's say for example the manufacturer wants to implement data or AI, uh, um, to find root cause behind certain things that are going wrong in the factory. If your practices from a root cause analysis perspective are very poor, or let's say you're um, and I'm talking all the way to Brightman, manager level or CEO level if your problem solving technique or problem solving methodologies don't work or they're non existent, how the heck are you going to use dashboards and data to solve a particular problem? So those basic fundamentals, those foundations need to be in there. That's why the place to start is to do a proper assessment from a world class manufacturing perspective. And you spoke about it early on. Where are your base practices and your fundamental ways of doing things in the factory lacking? And then if that's the case, what do you need to fix so that you can actually use the data? Another example is if, if the uh, if the plant doesn't have very good um, routine tier one, tier two, tier three meetings and they're not discussing the right things, you throw data into that, you're going to create a mess, you're just going to create noise. So, so, so I always say, and, and you know, the reason I got interested in talking to you um, a couple of months ago was that if you look at all the technology players in the world, the AI, the data players, they talk about boosting operational excellence but do they really know what operational excellence really means and what boosting that really means? And it really talks about the fundamental stuff around practices and improving processes, cases. It's not just throwing technology in this silver bullet. Um, so I think that's where you start. And then, and then finally sorry, once, once you get into closing the gaps, that point that you made earlier on around goal alignment and cascading, um, the KPIs is always going to be absolutely important when you implement technology. So technology is not a silver bullet. It's not something you just letter into this, you're hun mentioned. It's not something that you just bolt on. It's got to be built into your strategic process.

Speaker A: And on top of that, just to add to that because I'm 100% aligned with you, um, uh, the first comment I want to make is it's probably not always correct for somebody to when they want to go the data AI route to just throw money at, at analytics because they didn't do what you just said.

Speaker B: Yeah.

Speaker A: Understand where the problem is, where you, you start, what's the right things that you have to uh, understand. Because in our mind um, and like uh, when we talk about doing a joint assessment, um you look at certain things in that business to understand exactly where they are. And we would be looking at things like um, how stable is the plant running? So do they have good standardized production practices? How reliable is the plant? So what if they talk about oee? What is the OE numbers at? How much money are we actually using? So if we walk at, at the end of that assessment we should be able to give a pretty good return on investment number that if you going to implement best practices and information. Here's the numbers we're going to be chasing and that's the return on investment that we're looking after. And then to make that sustainable things like the goal alignment process would be crucial to get everybody on board. But there's many other elements that we can put talk about as part of that life. For instance, if you want to do go alignment you have to have the team structured correctly to do it. Like uh, because a lot of people think they've got teams because we've got production teams and we've got engineering teams and I've got a quality team. But they all three work together to deliver the output. So the real team is the operators plus the horizons plus the quality guy. So let's structure the team, let's align the KPIs, then talk about when are we going to make what information available and how are we going to pass onto them. So I guess what I'm saying, what I'm adding to your explanation is the assessment should lead to a uh, customized plan for that business that starts here because that's how mature you are. So first do this, get the change management plus how do we do the goal alignment? Um, if that's necessary, do we have teams to do that with? What is your state of technology and your data? So at which point do we introduce, introduce sensors and at which point do we make what information to what levels? So it will be a customized plan for that client organization of exactly what those steps should look like. Is that fair to say?

Speaker C: Yeah.

Speaker B: So here's an example of not doing that as an example of what will happen if you don't do that. So one of what we, one of the things we provide the manufacturer the ability to do is follow their own control plans. And Johan mentioned it around, um, removing variability in the plot. So now a manufacturer follows the control plans and they comply 100% and they come back to us and they say, but we still, we're still getting scrap or defect. What's going on? You dig a little bit deeper and you find that their maintenance practices are poor. So, so, and that's why I'm saying you need this holistic approach. So if you identify, if we did the assessment, you would have identified the gap from a mansion perspective. And, and if you do the work around doing maintenance practices, um, and getting that right and then control against their uh, own recipes in tandem, then you're going to get a much better ROI result. So, so, so my point is you need that holistic assessment and that holistic implementation.

Speaker C: I just want to frame of a simple term. It's taking the different perspectives from this knowledge base we have represented here today to help prioritize the next action for the customer. And I think that's a very valuable thing because there's so many potential solutions, tool, whatever you want to call it out there. It's what's the most kind of efficient, effective next step for you to take in your current context. I think it's the value of giving the customer the right perspective on their current state because that leads to the next right decision, in my opinion.

Speaker A: Can I throw a question at you, um, to unpack maybe as part of this you mentioned earlier on your number one enemy is variance. A lot of people think that Six Sigma is supposed to solve that for you because Six Sigma addresses, addresses variability to a large extent. And I think over the years tools like Six Sigma became all things to all men. You know, if, if you've got losses in your business, just, just you know, train up a bunch of green belts and black belts and launch a whole bunch of projects and you're going to save 30 million rand or dollars. You know what is your view on um, that's what people have always understood as reduction availability. So looking for a solution. When you spoke about variance early on, how do we connect that understanding of reducing variability versus this information world?

Speaker C: Well I'll maybe start to say that I think the tool is still the same. Six Sigma is still a very valuable tool and the right way to do it. My actually change is just can you do Six Sigma in a much more effective way, in efficient way actually. Because a lot of the effort that goes into Six Sigma or the unsustainability of it is the effort that you have to put into it to keep it sustainable. And that's where many good initiatives falter at the end of the day is not, not that it's a bad initiative, it's just the effort associated with the initiative is its own downfall. So how can we leverage technology, data, information? Because the great thing technology gives us is the ability to automate some of the maintain processes that is collecting of data, that is processing of data, ie visualization of data. Uh, you don't have a smart person spending six hours making Excel report. Now you actually have a smart person that has a report that's pre populated and they can apply their critical thinking to the next challenge. And that's how I see these two worlds play together. Six Sigma still for me it's the principles I fully believe in, how implement them. It's just the sustainability that I think in my humble opinion is where it falls flat many times.

Speaker A: So can I ask you another question on that? Because what we've seen, seen also in the past a lot is when a company is losing money so they decide to go on a Six Sigma journey, let's say uh, to reduce waste out of the business. So the Six Sigma journey and eliminating those loss becomes the responsibility of the green belts and the black belts in the business. Typically the industrial engineers, they give that responsibility, train them up, go and find the money and get rid of those losses. So never favor or some of offer. It does not become the responsibility of the line manager to reduce losses in his section. It's somebody else's problem. I, I presume and that's my question. It's the same when we apply information. If, if we, if somebody buys data, profit in to bring visibility of information analytics and AI, if they make it the responsibility of someone out there, it's probably, probably going to have the same effect because you keep on talking about an integrated holistic approach. Does that mean that when you embark on a journey like this, you need to think where it complements your strategy and who's going to take ownership for the visibility of information? Can you elaborate on that, if you don't mind?

Speaker B: Yeah, I mean, yeah, so I've been reading the Toyota and your question reminded me me of, of one of the questions I think that GM asked, asked to or something and, and they said so how many industrial engineers do you have in your business? He went, we've got 2,500 operators. So we have 2,500 industrial engineers. So it's the same thing. Right? Um, um, and I think your question relates to the same point in that you, you need to think about again, it's going to go back exactly to what Johan said. What's your strategy, what are your processes, what are your people doing and which are the parts of the business that you need to improve? And if those are the parts of the business or the processes you need to improve, then make sure that those people are able to use data and AI and make sure that they're empowered to use data and AI. Uh, and it might mean just simply using the data, it might mean having a, uh, an AI, uh platform where they can build no code model, you know, no code models or build models in a no code fashion. But they need to be empowered to make those decisions and very quickly use the tools that they dispense to, to drive value. So, so I think this answers your question, is that you think about again the goal alignment, think about the strategy, think about the vision, think about the areas where you want, want to um, use data or AI. And Tian's point, it might not be AI, might just be data. And then think about, and you've often heard about this even in your world is what are the as is processes and what are the to be processes? So if the process is XYZ now, what does it look like tomorrow when the data and the AI is in and have you designed your organization structure to deal with the to be process? Um, so you need to go through that as well. Um, and then the other things you need to think about are, you know, so if you, if you implement data, sops are probably going to change as you, you know, we spoke about go alarm, um, the KPIs will probably change, the daily routines might change. So you need to think about that, you need to think how that is going to impact the operators and the people that actually implement those daily routines. So that people element and that's why I spoke about change management being so important earlier on and, and having teams like CCIG on board. Because change management in a factory is different to anywhere else. You've got the SOPs, you've got all of those things which you typically wouldn't find anywhere else. Um, so the people element and getting them ready is a massive component.

Speaker C: Just want to go around the people one way where I typically challenge a customer. Ah, since because you talked about the silos, this industrial engineers challenge, they sit in their office and they get frustrated with operations not taking their cue or their information. That's leadership. That's the soft skills orange you alluded to previously. If you don't have leadership or the soft skills to align people and their focus and goals, then as an organization, all the technology in the world, there's no probably not going to solve that problem for you. It might highlight it, but I mean if you're not inability or maturity to act on it, you're going to have a tough time. But what I also want to challenge the audience to is typically we look at KPIs like scrap production leaders. KPI must not be scrap production, it must be how do you empower your team to reduce scrap? It sounds like the same KPI, but it's not. Not because if you approach it from that mindset, you actually see opportunities where tools like technology and methods can actually get you to that. If your KPI is just reducing scrap, it becomes an almost reactive culture of reacting to whatever number is in front of us today. But if you start empowering people of that mindset now methodologies and tools and technology starts to take a bit of a different space. Your KPI or objective is to empower your workforce to improve towards a goal. It's not the goal itself. And I think we miss a beat. Uh, even in the previous places I worked, we all fall into that chat. We have this nice KPI because it's measurable. It's the empowerment towards the KPI where the real magic happens, in my opinion.

Speaker B: But is it? I mean that's your world, right? Uh, how do you create the capability to improve?

Speaker C: Exactly.

Speaker B: Uh, and that's another reason you've got to combine these two worlds. That capability for change and improvement needs to be there. So what Johan's talking about now, without that capability for change and wanting to improve, which is what you guys do, uh, how are you going to get value from data 100%.

Speaker A: So, so let's just confirm quickly what do we mean with bring the tools together as a kind of a Conclusion is that when we talk about capability, being ready for AI, um, it means that people can adopt it. Like you use the words meaningful information quite a few times. You use the word integrated approach quite a few times. We also use the word integrated approach from a different perspective. Like how do you integrate various practices, the supply chain practices with manufacturing practices, with the outbound side? How do you, how does a CEO, uh, or an MD have an integrated approach to where the business is at and what do I need to improve? So when we talk about um, how to create the right environment for this to thrive in, one of the key things we talk about is how do we get people to come to work, to not just come and work for a salary when they come, uh, to when we get them, when we engage them in such a way that. Right. Change management, that they come to work because they understand what the business is there for, what are we going to achieve and my role in it. I'm a cleaner, with respect, I'm a cleaner on the shop floor and I know if I don't keep this area clean or this machine clean, the operators cannot achieve those targets. If I get the teams to move to that environment, they are so much more ready for anything you throw at them. Because so often, and I want to use this AI journey with, um, in comparison to health and safety so often. We've often been asked to help companies to improve their health and safety culture. But health and safety culture in itself means that your behavior has changed. People look out for each other, they look out for each other's back. They're always aware that it's not just there's our values, there's our safety rules, but we still hurt people. Because you've changed the environment in such a way that people are now receiving, receptive because they understand the change management is down. Right. They understand what they're there for. Now we can tell them, listen, the way we're going to have to improve safety, health and safety is this, this and that and they absorb it. And I guess this is exactly the same. Yes. And they adoptive, they're adaptive, you know, so when we prepare an organization to adopt information and AI in the right way, you need to make the, the ground right, the environment right for them to adopt that. So can we quickly, as a, maybe as a conclusion, just talk about, somebody might be listening here, uh, um, saying either they think that they're pretty good or they are good, but they also understand that maybe they don't have the visibility of information or they have done some extent of Data analytics, but they don't know where AI is supposed to take them. Or it might be an organization where they are struggling with performance performance. They're either not mature or you know, they're not performing very well. So just to conclude the right approach, what we've discussed in our view, data, profit and CSI growth come working together is to first understand where the business is at, understand how mature the practices are, how good is your production environment, maintenance M or whatever the businesses it might not be manufacturing, what is the condition of your data, what's the sources, how has it been used and then what, what is the losses and waste that we identify in that assessment? To understand why would you even embark on this journey of best practices information, the whole excellence journey? Why would you. And uh, then the conclusion should be there's a definite return. Either it is improved performance or it's more sustainable performance where we are at. But that would be the objective. Am I correct? Are there any other views on that? Are you comfortable with that?

Speaker C: I think that's an excellent summary.

Speaker B: Yeah, I go with that. I mean so just to paint the picture slightly differently, I think the risks of not combining the two worlds, of not doing that might give you some insight into terms of how to do it. So, so I think the first one is often companies uh, will, will look at AI and data and they'll try and find a use case, uh, let's reduce scrap in that part of the process. The problem with that is, is you're implementing what we call a point solution and it's not sustainable. So you might be missing systemic issues in the business which I think the combined assessment approach will uncover. And so your, your data and AI strategy needs to be geared to addressing the systemic issues and not the point solution. That's, that's point number one. Then I think um, if you look at the combined approach, it also looks at alignment with your goals and your strategy. Um, and so that's very important as a place to start. And how do you get goal alignment and KPIs trickled down to the rest of the business? And, and so the risk I'm pointing out here is if you implement technology, how do you know that you actually driving to the improvement of a particular KPI, that's a strategic priority. I um, think also if you just implement the technology. So again going down to where you start, you might be uncovering, and I gave example about root cause analysis, you might be uncovering patterns which talk to the symptoms in your processes but if you don't have the Best practices like really cause cause analysis, you might not get to the root cause. Then there's the human element. Um, so I think if you, if you apply just data or technology on its own, um, what's going to happen is you're going to disempower your staff, you're probably going to get them to disengage. Um, and that's why as a starting point you need this holistic approach. And then, and then finally I think if you implement technology without this holistic approach as a starting point is your solution that you put down is not going to be sustainable. Um, and the reason I say that is because of all these, all, all these best practices like good 5s and, and all of that stuff. The end goal is always sustainable capability. So, so you want that sustainable performance. You want to improve your KPIs tomorrow. And know often you hunt talks to customers about this. You changed your recipe. How do you know that tomorrow you've got access to that recipe? Ah, I don't know. You wrote it on a piece of paper.

Speaker A: Yeah.

Speaker B: It needs to be sustainable.

Speaker A: Sure.

Speaker B: So without that practice of recording what their recipes are and keeping them um, the, the application of data analyst unsustainable. So your ROI is unsustainable. So if you look at those risks that gives you so many indication of where to start. Bottom line is you've got to start with this holistic approach.

Speaker C: Great for me the journey, what makes it exciting again in metaphors, if you want to become fit, there's a lot of information you can have. But a great app like Strava, uh, helps you quantify your fitness journey. It makes you a lot more effective in your training. And this is the way customers must also approach this. It's you want come fit as organization. Why not get the Strava of the fitness world to you as well that you can track your progress through this journey. And that's where data, uh, and information plays such a key role because the Strava gives you a nice history of where you started at where you're going and why, where maybe there's opportunity for improvement. I you're not doing enough zone two training or whatever the case might be. And that just opens, just makes it such more uh, efficient and effective. Orange, you mentioned efficiency quite a few times. It's the same thing we all want achieve fitness but why not do it in a smarter way? It's a simple kind of challenge or sentiment to take from this.

Speaker A: So listen ar, thanks very much for your time. Johan.

Speaker B: Excellent.

Speaker A: Thanks very much. I think this has been an excellent discussion and I think for the audience out there, um, you know what? We, we read up a message articles currently, and you will see if you read globally, um, how people are struggling to create this picture, this integrated approach between best practices and excellence and this whole new world of information. And I really hope that we've created some clarity around this, how you deal with this, um, where does this emanate from and, uh, how expert companies like ourselves are going to tackle this problem and help companies achieve this integration. So thanks very much for listening.

Speaker B: Excellent discussion.

Speaker A: Thank you.