Key Lessons from PdM Success in Food & Beverage - a panel discussion

Trend Detection Podcast ·

Show notes

Welcome to the Trend Detection podcast, brought to you by Senseye Predictive Maintenance – the platform which enables predictive maintenance at scale across all of your assets, across all of your plants.

What you’ll learn in this episode: Join us as we explore how predictive maintenance (PdM) is transforming the food and beverage industry—with a special focus on the Middle East.

Why PdM is becoming essential for F&B manufacturers to stay competitive, improve uptime, reduce waste, and drive sustainability.

How the Middle East’s unique challenges—like water scarcity, reliance on global OEMs, and workforce turnover—make PdM a strategic necessity.

What a successful PdM rollout looks like, including a five-step methodology (Scope, Design, Deploy, Operate, Refine) that guides you from problem definition to scale.

Real-world stories of ROI, including how a global pet nutrition brand and a European dairy achieved fast, measurable wins with Senseye PdM.

What sets Senseye and Siemens apart, from AI-powered insights and multilingual support to full-stack, sensor-to-cloud solutions backed by local experts.

You can find out more about how Senseye Predictive Maintenance can reduce unplanned downtime and contribute towards improved sustainability within your manufacturing plants, by visiting: www.siemens.com/senseye-predictive-maintenance

Transcript

For our customers though, the benefits that they receive really is the maintenance of uptime, you know, having that early indication that there are signs of degradation so that uh with enough detail that a maintenance team can go and start to repair and fix what is going wrong. But they also benefit from precise and correct maintenance. So for the other 95% of their assets, machines, their operations that are in perfectly good order, that you have a centralized visibility of this. Um, it also helps and does touch on sustainability because, uh, or has an impact on sustainability because there is less time spent idling, less consumables used while performing maintenance, less emergency transportation of spares. So really predictive maintenance in an industry like food and beverage where it's high volumes, you know, you can sell as much as you can produce. Welcome to the trend detection podcast, powered by Siemens. Hello and welcome everyone. Uh, I'm James, and I'll be your host for today's webinar on predictive maintenance in the food and beverage industry. Well, thank you for joining us. Uh, we're really excited about today's session because predictive maintenance is fast becoming a competitive advantage for manufacturers, um, especially in an industry as fast moving as F&B. So today we'll take a closer look at what predictive maintenance is and how it differs from traditional monitoring. We'll look into how leading manufacturers around the world are already seeing real value in predictive maintenance. And of course we'll see how this is being applied within the Middle East market. Now if you have any questions, uh, please use the Q&A function that we have and we will ask our speakers your questions at the end of the session. Now, talking about our experts, I'd like to introduce you to Jonathan. He is the head of solution engineering for Sensi Predictive maintenance. And we also have Rakesh, our other expert, who is the digital enterprise services port portfolio manager here in the Middle East. Thank you both for your time today. Uh, looking forward towards our discussions, um, and perhaps maybe to kick off, Jonathan, if I could come to you and ask, maybe if you could just give a quick introduction onto predictive maintenance, um, especially in regards to the F&B market. Absolutely and good afternoon to everyone online. So predictive maintenance. Essentially, it's an evolution from traditional condition monitoring, but instead of having a specialist go around the facility taking manual measurements, then uploading them, uh, to the computer, analyzing it and reporting back on it, you know, I think you could think of that process like a vibration analyst, uh, going around and doing a manual process. This is now automated. So the process of data collection, analysis and reporting is automated. Uh, using technology, uh, and, uh. Technology and systems and solutions like Senseiie within Siemens, but like any good reliability engineer, the system continues to learn. It learns about failure modes, failure patterns, and starts to become very specific when it sees signs of degradation. But it's also not limited to rotating equipment and doesn't just use one form or one stream of data like vibration. It incorporates a wide range of data sources and data types to get a complete view of an asset's condition, signs of degradation, um, and really start to highlight those important details when needed. For our customers though, the benefits that they receive really is the maintenance of uptime, you know, having that early indication that there are signs of degradation so that uh with enough detail that a maintenance team can go and start to repair and fix what is going wrong. But they also benefit from precise and correct maintenance. So for the other 95% of their assets, machines, their operations that are in perfectly good order, that you have a centralized visibility of this. Um, it also helps and does touch on sustainability because, uh, or has an impact on sustainability because there is less time spent idling, less consumables used while performing maintenance, less emergency transportation of spares. So really predictive maintenance in an industry like food and beverage where it's high volumes, you know, you can sell as much as you can produce, that having the ability to Have a really good maintenance strategy in place, backed by a centralized visibility, um, it means that you can produce more and essentially make more. Right. Thank you for that overview there. Um, maybe Rakesh, you can give us a little bit of insight of like what Jonathan says, but what the relevance is then for the Middle East market. Yeah, thanks, James, and uh very good afternoon to one and all. I mean, the Middle East region is very special, yeah, I mean, It must address food security challenges, and this is resulting from water scarcity, climate change, hot climate, and regional instability. And food and beverage plays a very important role in uh the Middle East diversified industrial sector. Uh, the government of, uh, I mean, you can see regional countries like, uh, the United Arab Emirates and Saudi Arabia bringing initiatives such as, uh, the Smart manufacturing incentive program in, in Abu Dhabi that incentivizes sectors such as food and Bev, uh, to embrace Industry 4.0. So in short, yes, it does play a very important role in the food security aspect, uh, in the Middle East. Right. And then maybe while we're talking about the region, could you go into some of the challenges, uh, in some of the traditional maintenance approaches that you're seeing? Yeah, traditional maintenance procedures, I mean, uh, they can be less efficient, uh, resulting in increased downtime, higher repair costs, uh, greater reliance on manual processes, uh, and in the long run this is, uh, I mean this is leading to compromised quality without any predictive insights. And if you look at the food and beverage factories in the Middle East, they suffer from one thing is the heavy reliance on global OEMs. All of them are from Central Europe, East Asia, etc. So there's a lack of documentation of their plants. Uh, if you understand the dynamics of food and death plants, they're built by multiple OEMs over a period of time, and this brings in a lot of complexity with SOPs in operation and maintenance. And not to mention this, the region is driven by expatriate population, so there's higher employee attrition rate in the Middle East, and this can cripple productivity, erode institutional knowledge. Uh, it could drain resources, and this is prevalent due to the, the large landscape of end users. We're talking about 5000 F&B end users in the market with a revenue of more than a million dollars, yeah. So, these are regional specific challenges, James and Jonathan. So uh I would really say that over a period of time, uh, we need to look at regional specific challenges against global challenges, and probably technologies like PDM can actually fill that gap. OK. Well, on that one, the segue to Jonathan, like, so how can, how can like a PDM, you know, predictive maintenance enhance operational efficiencies and, and consistencies? Yeah, uh, as I mentioned earlier, that centralized visibility of your production, of your machines, your operations is fantastic. Um, and Rakesh, you, you said that having a reliance on multiple OEMs, a facility is created with a full stack of OEMs trying to merge these things together. Um, quite often they don't talk to each other, so you don't, they are siloed systems that you have to Somehow configure and make work together, um, in a way, or you don't configure them and you don't have that, uh, centralized visibility. But when you do start to incorporate that one central point that can speak to all the OEMs where you can have the visibility to all these operations, you are in control. You are, you have the The know-how now, the, this visibility to plan in your maintenance correctly, to plan in your production cycles, to have an understanding and better projections of when you're making, what you're making, how you're making it, and a reliability of these operations so that you know what you're making is of the highest quality. So really it is about gaining an understanding from all areas using your OT uh information, your IT information, and also the human element to bring it all into 11 point, um, so that you're in better control. OK, I mean, it makes sense. And then Rakesh, like maybe if you just bring it down a level lower, how does this really like tie in with like the Siemens portfolio, um, how can it help with like a customer's digitalization journey? I mean, in fact, Siemenstrial automation I mean we we brought a sensitive cloud solution and you know I'm talking about uh connectivity across levels. I'm talking about horizontal connectivity, vertical connectivity. And being a player in that industry, we provide that one-stop shop, uh, to provide an end to end uh turnkey solution for PDM offerings. Charlie should say that uh we one of the very few vendors in the world that offer this entire ecosystem uh from sensor to cloud, and uh it plays a very important role in the bigger accelerator uh initiative of Siemens, uh, bringing in an ecosystem of like-minded partners globally to actually collaborate uh and uh communicate between our systems uh very efficiently. And I'll just add on to that, Rakesh, Siemens also bring the, the specialists and the expertise in all of those layers and in the industries, you know, it's not just about the, the technology that Siemens provides, but it is the, the human elements as well backing the technology. Absolutely. I mean, we speak the OT language and uh that differentiates from the IT landscape and we, we provide the bridge between the OT and the IT world, and I think that's the perfect synergy for solutions like PTM. Yeah, definitely. And Jonathan, I mean, you, you're working with customers kind of all around the world, you know, do you have maybe any examples that you can bring, you know, from, especially from the F&B market? Yeah, absolutely, um. We've got a number of different types of FMB customers that we're working with globally. Um, I think probably one of the, the most interesting ones that are on my plate at the moment, that we have a customer that, uh, deal with pet nutrition, essentially pet care. And they've gone for a global deployment, so we're looking at many, many factories around the world. They have identified that in various areas, they have different requirements. So in North America, for example, they have a, a specific requirement for handholding, very much a, a managed service, so that is, uh, full investigations, sensor selections, installations, integrations with their existing systems. Um, that is training for personnel, that is even, um, managed service of looking at the results, interpreting them, and, you know, presenting that to the maintenance teams with recommendations. Uh, in other regions, if we look at Asia, they are looking for not a necessarily a center-led deployment but looking to really utilize their existing OT data and their, their existing investments. Um, and using or letting the technology and our generative AI capabilities be the, that knowledge and recommendation basis for them. You know, so that is a, a very interesting way to see how we have to approach the, the initial setup with them differently in different regions. However, the underlying technology is the same. You know, the results, the, the, the logic, the machine learning factor of it is exactly the same. It's just how we would implement that. But there are huge variants in the type of machines we're looking at for this customer. You know, some are really focusing on support assets, fans, compressors, you know, your utilities. Others are heavily focused on their main production assets, your extruders, chocolate conches, and so on. Um, uh, oh yeah, those types of machines. So very varying approaches to implementation. However, when we look at the, the value of this, it really is around maintenance efficiency. So reducing your reactive, your unplanned events, uh, making sure they have better visibility of their planned maintenance approaches and really moving them into a, a more efficient way. If we look at some of the other successes from some of our other food and beverage customers, there's um a dairy in Europe that we went and investigated. Um, I was there on site personally. And what we found is that in their existing infrastructure, they had very little control. Their systems were extremely siloed. They weren't sure whether or not we could get access to the data which was there, so we had to plan for a, a hardware-led deployment. Anyway, long story short is, we came in, we made the recommendations, we helped to make sure that it was installed correctly, um, that various sensors were put on the right components in the right way and data was coming through, and they, they got a return of investment for their entire investment including hardware with one, with one save. One notification, um, on a very critical piece of equipment meant that they had a return on investment immediately. So a very short space of time, but it is about targeting the value in an environment and uh we have a very structured process which I I'll, I believe we'll get into it a little bit later. Yeah, I mean, you're talking about like, again, we've talked about the efficiencies and so forth, but the other side of the coin which a lot of our customers are asking about is around like quality control, um, and, and like, for example, compliance, like these types of legal issues. I mean, where can like predictive maintenance come in there to, to support? Yeah, absolutely. Well, with a better or a more reliable plant or a more reliable line even, you're going to have better product quality. So if you know that your components are running in a perfect condition or in a very good condition, that you aren't having tolerance drifts, uh, that you have The components are running the way that they should, you're going to have better quality. So some examples of this. is we have been monitoring for one of our customers exit gauge. So product goes in through rollers to get a desired width of product coming out before it goes into the cutting then packaging process and by monitoring the exit gauge, a, a quality parameter, we've been able to see where some of these rollers start to become deformed or the tolerance of the, the rollers starts to move away. So it shows us, it gives us an indication of mechanical degradation. However, by spotting that early, we can then make sure that they aren't doing a run. Excuse me. There is a fire alarm test this morning. Well, at least we know it's live, yeah. Yes, absolutely. Um, yeah, so by monitoring this quality parameter, we can make sure that we give them early indication that there is something wrong from a mechanical degradation perspective, but also that they haven't spent 2 days or 2 weeks even producing and packaging and then starting to distribute uh products which have got a bad quality behind them. Also, if you look at various things now looking at mechanical uh parameters such as if you look at a sealer bar, a sealer bar is essentially an element. If that sealer bar is not running correctly or is not operating correctly, let's say it doesn't have a uniform, um, heating profile across the bar. When that goes to seal your package, that could then um leave spaces open on one of the sides. So that has a huge product quality and then can, uh, product quality impact. Which probably won't get picked up or may not get picked up immediately, um, and therefore you have recalls because packages aren't sealed, you know, um, food products are now exposed to whatever elements they have been and there is a problem with that. So again, just recap that, by having better reliability, you make sure that your standards and processes are followed better and that you have less potential for issues later on. It makes sense. I mean, OK, this is, this is all good. I think there'll be some people in the audience who perhaps haven't, uh, deployed predictive maintenance yet, you know, like, so where would you, for these people, where would you, what would you recommend? Like what are the first steps and like the best practices to get started when it comes to, uh, your first, your first introduction to predict. I always say that any one of our customers needs to start with the problem definition, defining what it is they're trying to solve. If they're just wanting to implement technology for the sake of it, that's probably not the right driver. But if you can or if you identify in your environment that you have maintenance inefficiencies, That your teams are overworked, um, that they are having to firefight the entire time, that the cost of downtime and this lack of reliability is huge, that you have potentials for increasing your OEE, um, and your output, then that is a good place to say, yes, I want to start. So you have a problem definition. And then for us, our structured methodology, I say it's quite simple, it's quite simple for us, but we guide our customers through this. So for us, we, uh, we have 5 phases in every single project. These are scope, design, deploy, operate, and refine. And within each of these phases, there are 4 steps. But again, as I said, we manage the process with our customers. So if I just uh describe them briefly. Scope phase is really information gathering within the sales cycle. We make sure that there is the correct problem definition, that we are looking at solving problems and really giving or generating value in a customer's environment. Then we would assess the assets, where this value is gonna come from, the data, infrastructure, who's gonna be involved in the project so that Um, we get a project team together and they are aware of their inputs, their requirements, and we define the success criteria, all within the sales cycle. Then once the project starts, we go into the design phase, which is an in-depth look or getting into the nitty-gritty of the project. You know, operational interdependency between machines, um, you know, what are the actual communication protocols between various infrastructure or architecture systems. And so on. Then deploying is about making the data move. So if there is need for additional hardware that that gets installed, that systems are integrated, as um data is streaming into the application, assets are on boarded, users are on boarded and then trained, which typically takes about 1 to 2 hours. It's very, very It's an intuitive system and it doesn't require that much training, thankfully. The operate phase runs through a best practice approach on using the application, monitoring notifications, adding in work events, doing daily, weekly, monthly reviews, and then in the refined phase, it's about making it as good as it can possibly be in that environment. So we assess the need for additional data if necessary or potentially higher processing of data. We look at who's using it and the adoption factor, you know, it doesn't matter how good your system is, if it's not being used, it's a, it's not a good thing. So we look at who's using the application and making sure that we are getting the adoption from the users. Um, we also look at the success that's been achieved. And we have a process of documenting these success because what we find is in any project, you'll have your project team. And, uh, the project team may know exactly what's happening, but they need to be the internal, um, uh, success distribution channel. They need to make sure that they are distributing the wins internally so that they can get wider buy-in and they can look at scaling the project, um, and having it successful throughout an entire plant. …

Full transcript (WebVTT)