The promising future of AI tech in Life Sciences – Transcript
To see the show notes on this AI tech in Life Sciences podcast, click here.
John Nixon: Welcome to the Industry Forward Podcast! Here we explore the key trends, transformative technologies and real-world innovations that are reshaping fields from aerospace to energy and beyond.
I’m, John Nixon, Global Vice President of Process Industries at Siemens Digital Industries Software. I’ll be your host today.
My guest is Patrick Ansems, the Global Head of Life Sciences at Siemens Digital Industries Software.
And we’re going to discuss comprehensive digital twins and artificial intelligence (AI) in the Life Sciences industry. Mainly what these tools will look like and how they will help us both in the present and the future.
Pat, can you share a few specific real-world examples [of] Siemens technology (whether it’s digital twins, or AI, or the digital thread throughout the life cycle). Can you share with us what has delivered measurable impact in the Life Sciences industry?
For example, how have companies used these capabilities to improve quality, reduce release times, or accelerate the transition from development to manufacturing?
What made these digital capabilities so critical to achieving those outcomes?
Patrick Ansems: Yeah, I’m more than happy to. Because I think the most interesting conversations are always how our customers [utilize] the tools that we provide.
So, one example that comes directly into my mind is when we talk about Terumo Americas, which is, let’s say, a global leader in medical devices. They had the initiatives to move away from paper-based manufacturing into quality processes and build a much more connected and digital factory environment.
What they have been doing is by digitizing the manufacturing execution and quality workflows, they’ve been able to take, let’s say, material batch release from a full day of work to under 30 minutes. So that is, let’s say, a significant reduction in the amount of time that they spend on paperwork. And therefore, they’ve been able to reduce [nonconformance] and complaints by about 40 to 60%.
And the other example that comes to heart is a customer that I enjoy working with, which is Johnson & Johnson Innovative Medicine, where they’re kind of taking their digital process twins and they are being utilized to optimize the active pharmaceutical ingredient production. And instead of relying on physical trial and error, The teams can now simulate and test and refine parts of the process virtually before we start applying them into production.
So, in one of the examples that they’ve published, that helped reduce the solvent switch time by 30% and the overall cost by around 35%. And those are two examples out of manufacturing, but also, let’s say, in R&D.
We’re now seeing some of these initiatives where they’re utilizing that contextualized data and the digital twin process to build – almost like the mind of a scientist within R&D. Where they’re not just storing experimental data, but they also understand the context around that [data]. Leading to, let’s say, substantial reduction in the amount of experimentation that they need to do. Sometimes even up to 40% of reduction in experiments. And subsequently now actually having the information to start to do a much more shortened development cycle and bringing those drugs to markets in a much quicker way.
John Nixon: What does a comprehensive digital twin really look like from a Life Sciences industry perspective? We’ve talked about it again and again.
But, I mean, I want to talk about what roles [it will] play? And how does this feed into AI, as we’ve been discussing.
Patrick Ansems: Yeah, so to me, the comprehensive digital twin means that you can build a virtual model of the product, the process, and in the end, [the] patient. By providing, let’s say, digital capabilities that allow you to match that to the real-world evidence coming out of your manufacturing facility or coming out of your R&D lab or from wearables or real-world evidence around the patient. That’s when you start to provide the most … impact that you can think of.
Look at it this way, by combining the physical and the virtual world, you can start to do a lot more, let’s say, simulation-based analysis rather than verifying or building out experimentations within the laboratories or within your manufacturing facility.
And when you start to combine that as a, let’s say, solar data fabric and you start to apply to that, what I always call a composable architecture layer. So, a layer on top of that will allow you to build a knowledge graph around that and analysis and agentic [AI tech] on top. You start to build that really what I would consider that pharma-in-the-loop concept.
So, you’ve been able to match the information from the virtual world from the conceptualization of an idea of a molecule all the way of how that molecule is being manufactured. And you can align that with what’s happening in the physical world and start to apply agentic AI. So, you can actually start to utilize [all] these AI algorithms and tools to design and [manufacture] an end-to-end research and manufacturing process.
That’s really that pharma-in-the-loop concept. So, having that iterative loop where people in R&D can learn of what’s happened in the manufacturing downstream, like are there certain materials that we should not be using because they are hard to source or are expensive in the manufacturing process? Or are there conditions that we have tested in R&D that could have a similar type of effect in the manufacturing process?
Like [all] that information, being able to correlate that I think will have, let’s say, a tremendous impact to the field of Life Sciences.
John Nixon: And as you’re talking about that, the image that comes to my mind is: I [think] you’ve got the nuclei of atoms, you have atoms, you have molecules leading to compounds, leading to materials. It moves out of the lab. It now moves into scale up. And then we move into manufacturing. There’s this provenance of data that is so important.
And to me, when I think “comprehensive digital twin,” and “the human condition that we have to apply it to,” [there is] importance of that provenance. [There is] importance of that traceability to ensure that what we call a comprehensive digital twin truly is comprehensive. It is that entire journey to that one human being. And if you have that provenance, … that’s well orchestrated, traceable, secure and trusted. Now we get into AI.
And that’s what I want to ask you next. What will [AI tech] look like from a Life Sciences industry perspective? And you’re starting to discuss the role it will play. Take me down that path.
Patrick Ansems: Yeah, so, I mean, a couple of examples that you can already see, [set] as the applicability of [AI tech] in certain, let’s say, specific use cases. We have already examples of protein folding and protein binding, right? So those are really … well-established domains where [AI tech] is being utilized.
What AI doesn’t do yet is model that entire end-to-end process. But as mentioned, I don’t think that that’s that far away, right? So, we now have access to [all] these tools, we have access to the models, we have access to [all] the types of information that we would need [to] build out a true digital twin of the process of the product as well as of the person.
So now it’s more of: can we [scale] the utilization of those models to that extreme across that entire paradigm? And I think we’re not that far away. We have all the tools in place to [start] to do something meaningful and impactful.
It’s just a matter of can we get all the contextualization of that data right [to] start to apply it through, kind of, end-to-end AI into that context.
John Nixon: Yeah, I hear you on the trusted path to context.
And so oftentimes, we have these conversations. And the first question that people often have is, “where am I today?” And “how do I start from where I’m at today going forward?”
I mean, here’s the question. How do Life Sciences organizations, how do they get themselves digital twin and AI ready?
I mean, how can we (you and I and our collective team here) how can Siemens help the industry digitalize from the lab, from that design effort, scale it up to manufacturing and then on to the, ultimately, the customer we all serve, the goal we’re all aiming for, which is that individual patient?
Patrick Ansems: Yeah, so, when I talk to customers about “where do we actually start, right? I mean, it sounds all very appealing, but how do I actually get it into my organization?” I already explained, or I would like to explain to them is that they should think big, right? You should think [of] great ideas, right?
Because [of] that, let’s say, that light bulb moment, if you start thinking about what is actually possible, will give you the greatest sets of ideas and that will allow you to think a little bit more outside of the box than what people normally do. Because they will look at their own specific situation within their own specific group or department and they will look at something where they think [AI tech] can provide something tangible.
So, the first objective that we have as an organization when we talk to our customers is for them to start thinking outside of that box and start to think in big ideas. What you then start to do is take that [big] idea that they have and digest that down into consumable blocks. Because that big goal is something that is not always directly reachable, because you need a lot of information, that information is not contextualized, [all] the things that we talked about [in a previous podcast]. But take that big idea and break it down into sizable bites and then work on a project with those customers in those sizable bits.
We’ve done work with customers in manufacturing and what we’ve done together with Roche and with Accenture and [with] NVIDIA. We’ve built, let’s say, a comprehensive digital twin of that manufacturing process. We’ve done trials where we’ve worked with customers more in the R&D space that they are now utilizing [AI tech] to generate the next iteration of bispecific antibodies. And utilize that information to determine how likely it is that it will be introduced or [is] being produced by a cell line.
So, think big, digest it down to a sizable bite that you can digest and that you have access to the data and the people that can help do something meaningful. And then the next comment I always [say], is “just execute.”
I mean, we as a company are always willing to do all these kinds of co-development projects. Where we’re taking these ideas and we’re building out a prototype, and we’ll test that for customers. We’ve done a number of iterations around that in our industry, in the Life Sciences industry, but also in other industries. We’ve done some fantastic work with PepsiCo around the [Siemens Digital Twin Composer].
And I think those are, let’s say, great initiatives that we already have that you can easily start to apply to other industries. So, think big, get it down in bite-sized chunks, talk to us or our partners around some of these initiatives that are tangible and execute.
And I’m always inviting our customers, or people that listen to this podcast, to kind of reach out to us and say, “hey, I have this great idea, but I don’t know how to get there.” We as an organization have the tools, the partners and the capabilities to make something – that idea come to life. And when you start showcasing the art of the possible to customers, we start to really kind of explore what other capabilities they can leverage to build out a digital twin.
John Nixon: You know you mentioned PepsiCo and the art of the possible. The exciting thing is in sister industries to Life Sciences, it’s moved from the art of the possible to what’s now becoming the everyday expectation, where we can have a physics-informed environment at scale.
With that said, Patrick thanks for joining us today.
One theme that really stood out from our conversation is that the future of Life Sciences depends on connecting data processes, people and patients through a comprehensive digital twin.
You know, as [AI tech] continues to evolve, success will come down to having trusted, contextualized data across the entire lifecycle. And that’s from discovery and development to manufacturing and patient outcomes.
And for organizations wondering where to start, the advice is simple:
- Think big,
- Break it into manageable steps, and
- Execute
Thanks to everyone for listening to the Industry Forward Podcast. Be sure to join us next time as we explore the trends and technologies shaping the future of industry.
And until then, I’m John Nixon. Keep learning, keep innovating and keep finding new ways to process your process industry knowledge.
Thanks for listening.
To learn more about AI in the Life Sciences industry, click here.
John Nixon – Global Vice President of Process Industries at Siemens Digital Industries Software
As Group Vice President for Process Industries at Siemens, John leads a global team that helps process industries leverage digital solutions that enhance efficiency, accelerate innovation and achieve sustainability goals.
John has over three decades of experience in strategy, operations and technology deployment for energy, chemicals, Life Sciences and CPG. He is well versed in the operational and business pressures of industry, including regulatory demands, decarbonization, talent gaps and the push for innovation.
Patrick Ansems – Global Head of Life Sciences at Siemens Digital Industries Software
As Global Head of Life Sciences at Siemens, Patrick helps pharmaceutical and medical device organizations build comprehensive Digital Twin technology that spans development to manufacturing, and beyond. These tools accelerate innovation and deliver better therapies to patients.
Patrick has 25 years of experience ranging from wet labs to digital transformation. He understands the complexity of scientific workflows and how to translate them into strategies that deliver business outcomes.
Connect with Patrick on LinkedIn
All trademarks are property of their respective owners.
