Scaling Data Foundations, AI for Future Pharma Manufacturing
Pharmaceutical and life sciences companies are exploring advanced digital technologies to help speed up development and manufacturing. The Digital Twin, Industrial AI and the Industrial Metaverse can offer solutions, but operate best when built on a strong data foundation.
In this episode of the Future Ready Podcast from Siemens, hear experts on the pharmaceutical and life sciences industry explore the potential of Industrial AI and the Industrial Metaverse, as well as the importance of strong data foundations.
Listen to the podcast or read a transcript of the conversation below!
00:00:12 Conor Peick: Hello and welcome back to the Future Ready Podcast from Siemens. I’m Conor Peick, a marketing writer at Siemens and one of a few hosts that you’ll find on the Future Ready Podcast feed. Today we’re bringing you the final part of a discussion with Maria Grom and Andy Whitehawk on digitalization in the pharmaceutical and life sciences industry. In this episode, we’re going to continue to break down the impacts of AI on drug discovery and production, look ahead to the industrial metaverse and how this might change how companies use and aggregate data, and then wrap up with some final thoughts. Thanks so much for joining us and do hope that you enjoy the episode.
00:00:47 Maria Grahm: And I would like to mention also some, maybe some AI applications that are maybe not so obvious. And that is how, because artificial intelligence is also great for visual analysis. Andy and I recently visited, we were at an event in Scotland where doctors were presenting on how they use the camera, the GPs in UK are now being able to use a camera, your mobile camera to do a first analysis on like maybe a patient coming in and wanting to have a look at a birthmark or something. And with the help of AI, they can more quickly draw the, you know, is there something dangerous or not? So that the using AI for visual analysis is something that can be used. And one of our customers in Denmark that we’re working with have lots of inspection cameras in their production areas. And they have come up with all these use cases of where they could use the data that they have acquired through these cameras to analyze and to be able to detect, for instance, if the operators are behaving in a way that there is a risk to create deviations—if the operator has a, there’s a risk of contamination, et cetera. And then with the help of this analysis of the camera data, they can then inform the operators that, oh, actually, maybe you need now to go and wash your hands extra because you maybe weren’t aware of it, but there was a risk of contamination. So this was one use case. And also this camera data has been used for example in the packaging, if there is a, how do you say, a label or something that sits wrong, or if something is not packed in the right way, the footage and the analysis from AI can detect these anomalies so much faster than a human being, where they have in the past have had people doing this control, and now we can use AI with the help of this visual data. So it’s not just within the actual drug and the production, it’s also in the secondary part where we can use AI.
00:03:24 Andy Whytock: I really like in that second example there around the AI cameras or CCTV cameras that are basically watching the operators doing their job and making sure that they’re doing it properly, is that the unions were so on board with that, because it was actually not about controlling—well, it is controlling—but it’s about showing where and advising if things need to be done to make sure that we stay within the regulation. So helping them to do their job better. It’s not, you know, to pull people up or to discipline them, but to make things better. And I think that approach of using AI in a safe and a constructive way, that’s where we can see these impacts. I love that example. I mean, it’s a simple example, like Maria used there, and it really has brought value to the customer or to their process, and it’s brought value to the operators and the people themselves, you know, not just for replacing people. I think it’s a very powerful example. We have many others as well, of course, but we probably don’t have the hours and hours and hours that we need to talk through all of our different examples of how we’re seeing this in the industry and how we’re helping our customers on that.
00:04:29 Conor Peick: Yeah, no, I think it’s a great example. And it does make sense from an operator perspective, maybe even it’s sort of an additional assistant, right, to help you perform your job as best as you can. And yeah, that’s a very fascinating example. In general, it seems like what we’re looking for right now is ways to incorporate AI into structures and systems that essentially manage risk, right? It’s about not letting an AI just take over completely. That’s not what we want—to make sure that it’s incorporated in a way that we can monitor and manage and such that also we take best advantage of what the AI is good at. Another simple example I’ve just recently experienced is at a recent doctor’s appointment, my physician used an AI-based transcription device to capture better notes about our visit and the things that we talked about. So I imagine even something as simple as that, where that can help him take better notes—maybe he doesn’t catch everything I say, or when he goes to just write down his sort of debrief about the visit, maybe he doesn’t remember everything that we talked about. And so maybe there’s a detail that this system can pick up, and then it’s captured, and then we have better information to sort of guide my care in the future.
00:05:59 Andy Whytock: Absolutely. It may be out of place, but it’s well known that it’s almost impossible to read a doctor’s writing, yeah, when they try to write things down, you know? So I think that’s a very good use case. But I mean, again, we see this, it’s about bringing efficiency. It’s about finding those use cases, whether it’s in healthcare and application like you say there or what we were talking about with making operating better. There are many, many different use cases. And I think that the key, coming back to sort of the original question, is it’s not just about isolated use cases and pilots. It’s about seeing this as an overall approach. And I think that’s what’s really, really key. And of course, not everyone has that in their scope to figure out how a doctor is being better at taking his notes to how an operator is being better than how a drug is being—how the med—how all the different experiments are being pulled together. But it’s a comprehensive, you know, it’s an approach that we all know brings a lot of potential. And we’ve got to harness that. And we know that we know and our customers know that this AI revolution that we’re talking about is really, really key to being able to make everything so much better. But it has to be controlled and done in the right way. And that’s where the traditionally conservative aspects of the pharmaceutical and life sciences industry, who have patient safety as the number one thing that they need to be focused on, have to be sure that they implement this in the right way, which is why they need people who understand the industry, the people that understand the processes, software, expertise, and so on and so forth to use it in the right way, not just to, like I said earlier, to take shortcuts. So we see this. And I think it’s a fascinating, fascinating area in which we’ll see much, much more in the near future.
00:07:41 Conor Peick: Yeah, it’s an excellent point. So I think we’ve talked now about first step into the future, digital twin, right? And now we’ve talked a little bit about kind of the second step being maybe AI, figuring out ways to apply this technology in pilots and then more broadly. So then the kind of the final step I think that we see when we look into the future is this idea of the industrial metaverse. And what a lot of people, I think, are seeing when they hear that term or when they encounter media about it is these really immersive visualizations, excuse me, high-fidelity 3D visuals and immersive environments. But the other aspect that I think is really critical to the industrial metaverse that we try and talk more about is the idea of data aggregation, data contextualization, which we’ve talked a little bit about this so far with bringing in, trying to digitalize the data transfer, the tech transfer. But with the industrial metaverse, you’re talking about doing this across, you know, huge, complex global organizations, supply chains, maybe even managing partner relationships and things like that. So, Maria, when we look at the industrial metaverse as a technology, how do you think pharmaceutical companies will be using this new technology, the industrial metaverse, to evolve how they collaborate, how they make decisions, and how they manage their whole life cycle?
00:09:15 Maria Grahm: I would say that I think it’s like the next, like you say, it’s the next step of the digital twin, because the metaverse is for sure the fully, the total digital twin of everything that is represented there. I saw a really interesting presentation really from also from a pharma customer on more on the again on the secondary on the warehouse side on how they are using the use of that immersed description or visualization of their warehouse, and the people that are also in there and the AGVs, et cetera, and how they manage the workflows within that environment and how it can really help them. But coming back to your question on the collaboration side, because we also—one challenge that we didn’t bring up is the shortage of work labor. And here again, being able to really work across in this environment, teams across the globe, will also help. So this will improve the way of having these efficient engineering teams in the future as well, and their being able to be in your facility or in your environment and for both for, I mean, when you develop it, but also when you are up in an operating mode and you need to maybe to check something—you have a deviation somewhere and you need to go in and to check that. It’s more and more difficult to have people working in shifts, so to say, and there we need to come up with new ways of working as well for these teams. But, then we… again, we talked about what is a digital twin and we also talked about, so now metaverse. I would say from the building point of view, I know I have been in projects some, several years ago where we had this quite comprehensive model of the building itself and how it’s built up and that you could sort of enter this with the help of software, entered this building even before it was built. So I think it’s, again, maybe we also need to define exactly what do we mean with metaverse because these digital representations of the building, et cetera, have not, it’s nothing new, but maybe it’s just more of the, that we’re now using—we have more data power behind it with the GPUs, et cetera, to being able to also going into more detail on every level of the simulations of each and every aspect of it. But I believe that we talked also about the people aspect and how people change. As we get more and more into working in these environments, I think there will be benefits for already from in the early stages for the scientists, the engineers, and also the plan teams in all the different along the value chain. There will be no easier way to collaborate. As I said, we have silos today and while having this fully digital representation that the metaPERS is, that will for sure also improve the whole efficiency of what the pharma companies are able to bring to the market and also produce. So I believe maybe it’s not in the nearest future, but it’s definitely a direction that we are going into the metaverse. And we’ve seen some good examples. We have our Medicines Manufacturing Innovation Centre in Scotland, where we have a digital representation, and that is an innovation centre where we work as a technology partner at Siemens. And we have…
00:13:41 Andy Whytock: As I say, perhaps I can build up on that, because I was actually going to illustrate that. Maria, the digital twin that we built there, which was to very much understand the layout of the line and to identify bottlenecks in this brand new production line that they were doing now—nothing particularly groundbreaking there in some way of doing that through a simulated model of the plant. But what we did was then we extended that to build the full plant, to be able to really get the full layout of the plant. And then you move it onto these calibrated models, which can then be used for training purposes of people. And you take it an even a step further, you know? So, and then build on that. And then when they’re building out an extension, they’re just rejigging that model again through the big, starting with plant simulation, but then into that photorealistic resimulation. And that video is available on our website of how we’ve taken that from a plant simulation model into a metaverse version for, like I said, for production throughput, for better layout, for training, for basically just optimization. Yeah, I think that the other one, but.
00:14:50 Maria Grahm: Starting, like we said, as well, on being able to provide to reach that vision of the metaverse, you need to start with the data and having your data quality in order, so it’s a long journey before you’re there.
00:15:04 Andy Whytock: And I think we’re seeing, we’re seeing again, this is quite interesting actually how we’ve gone through, these different sort of phases and these of, digital twins and AI. And now we talk about the metaverse. And I don’t think that it’s just a future promise that’s somewhere away. We’re certainly seeing elements of that. And we’ve a customer now talking about how they’ve been able to build that plant for their brand new build in the US, somewhere specifically. So we’re seeing that. We’ve got a lot of experience as Siemens of doing that for our own factories, of building every single element of that. And there’s some great presentations out there on the web of other industries. The pharmaceutical industry, we still, and that’s a little bit the role of myself and Maria, is to go out and meet our customers and to talk to them about the potential that the metaverse can bring to them. And that’s still a, I wouldn’t say it’s a battle, but it’s still a challenge to go through that and to show and demonstrate how that can be done in our specific industry. And I think there’s a really an interesting challenge as well.
00:16:05 Conor Peick: Yeah, I think something that I was just thinking about, because Maria, you mentioned we are encountering a workforce changes, some shortages in skilled labor. And I think it’s an interesting thing to look at this combination of this trend in the workforce. You have emerging technologies like AI, even more autonomous production scenarios, I think we brought up earlier. And then eventually more industrial metaverse type applications. It’s interesting how you can see these things stitching together to sort of extend expertise across the globe, even if you don’t have necessarily enough human people to put them across the globe. Maybe you have fewer, they’re more centralized. But through these digital tools, maybe we were able to connect them even in an immersive environment, for example, connect them to people on the ground to guide in scenarios where their expertise may be more necessary. And then you also have AI tools and other things to help maybe manage day-to-day operations. I don’t know, it’s an interesting kind of projection you could see of how maybe this evolves. And it ties in, Andy, to what you were just saying, that how this is future state stuff, but there’s aspects of it. There’s threads that we can pull in to today and see how it’s already changing the way that we solve problems and try and optimize.
00:17:36 Maria Grahm: Co-pilots we are already using in various ways. We haven’t maybe mentioned that, but that is for sure also an AI application with a large language models where you have as an operator a great benefit already to be advised or as a programmer also on coding. So there are many examples that are already up and running that we have.
00:18:05 Conor Peick: Awesome. So as we wrap up, I’m wondering what you guys see as maybe some guiding principles that pharmaceutical leaders or any leader for that matter who’s listening to us, what guiding principles they might be able to take away or keep in mind as they navigate the increasing complexity and change of various industries today and moving into the future?
00:18:28 Maria Grahm: Would you like to start?
00:18:29 Andy Whytock: I can start. I think it’s a great question, of course. And I think that maybe I put it this way. Two or three years ago, you were hearing in the industry, and I’m thinking about industry-specific conferences around the potential of AI. We talked a little bit about that today. But in the last conferences I’ve been at, we’ve been talking about building out the data to be able to do that, having that understanding, modernizing the factory so that you can then be able to leverage these types of, that type of potential. And it’s not just AI, right? It could be digital twins, it could be the metaverse. So I think it’s something that hasn’t changed as a statement for 10, 15, 20 years since I’ve been in the industry is start with the basics. Digitalize, get your data right, bring it together, break down those silos, those data silos, take the data out of those data jails. It’s something along those lines, yeah, around really getting to grip with the basics of what you want to be able to do to be able to then go and leverage that full power. It’s not a short journey, yeah, but there are many different journeys and there are many different ways in which you can break that down. So don’t be overwhelmed by it. I think that’s the point. Think about how to break it down to be able to get to build your factory of the future, your digital enterprise of the future.
00:19:49 Maria Grahm: Yeah, and if I may build on that, I mean, first thing I would like to say, and I mentioned it before, that it’s not just about technology, it’s also about people and new ways of working. So for the leaders, pharma companies and others, it’s important to understand that maybe it requires a new way to lead your teams in the future. It’s also about getting into these topics yourself because this is going so fast. So you need to be on board and really understand what does it mean for me and for my company and how will this change maybe totally disrupt the industry which I am in. So you need to educate yourself on this. And then I would also like to say that you don’t need to be a techie to understand artificial intelligence. You can have any background and really use these models that are there. It’s really easy. You don’t need to have an engineering degree to approach this topic. So I would say just immerse yourself and learn as much as you can. And then of course, come and talk to us because we love this topic and we love to be part of really supporting in this journey for our customers as well.
00:21:16 Andy Whytock: And to build on that, how do you come and talk to us? I think the best way is we have a, there’s a landing page, www..siemens.com slash pharma. That’s P-H-A-R-M-A, of course. And that gives you some links to some of the things we’ve been talking about and also the ability to engage with through the contact forms with the people in our organization. And it doesn’t take very long for you to be able to get to us through that, but also through LinkedIn as well. Right, Maria, we’re very happy and we do engage with people through LinkedIn because we’re always, like we said earlier, very passionate to talk about this and very interested as well to hear about the challenges and perhaps even whether we’re right or we’re wrong. Yeah, always interested to have that discussion. Yeah.
00:22:07 Conor Peick: Thanks again for joining us on the Future Ready Podcast. If you enjoyed the discussion, I encourage you to subscribe to the feed, as we will have many more conversations with experts from various backgrounds, diving into the most important trends and technologies in industry today. So thanks once again for listening, and we hope that you’ll join us again soon.
Siemens Digital Industries Software helps organizations of all sizes digitally transform using software, hardware and services from the Siemens Xcelerator business platform. Siemens’ software and the comprehensive digital twin enable companies to optimize their design, engineering and manufacturing processes to turn today’s ideas into the sustainable products of the future. From chips to entire systems, from product to process, across all industries. Siemens Digital Industries Software – Accelerating transformation.
