Host Intro (Gene Kim)

Next up is a team from Siemens. Carsten Spies is head of R&D of teamplay, the digital ecosystem platform at Siemens Healthineers, and Peter Fassbinder is principal key expert at Siemens AG.

This is such an exciting presentation for me, partially because it is in the healthcare industry, an industry that we were seeking more experience reports from. They will describe their internal partnership to bring digital transformation to Siemens Healthcare, another Horizon Three initiative.

Carsten will describe how their work is enabling the creation of new capabilities for healthcare providers, leveraging the internal consulting capabilities of Peter's organization. Peter will also describe the Siemens career track for individual contributors and experts. Please welcome Carsten and Peter.

Dr. Peter Fassbinder

Thank you very much, Gene, for providing the opportunity to present here at DOES London this year. Thanks to the complete IT Revolution team for making this virtual event a reality.

Our talk today is about how we leverage continuous delivery and DevOps approaches to stay competitive in a highly regulated market. We have two presenters today. I will cover, in the first part, more of the general DevOps aspects across the company of Siemens. In the second part, Carsten will give you some insights into a concrete example in Healthineers.

I am a principal expert for PLM process innovation with Siemens Advanta Consulting. I have over 15 years hands-on experience within Agile, and for four years I have been one of the key drivers of the DevOps transformation within Siemens.

Carsten Spies

My name is Carsten Spies. I am responsible for the R&D department of teamplay at Siemens Healthineers, and I have around 20 years of experience in software development.

Dr. Peter Fassbinder

Why is a company like Siemens adopting continuous delivery and DevOps approaches? As you probably know, we are a very large company. We have 385,000 employees today and revenue of 87 billion euros. We are structured into five companies: Smart Infrastructure, Digital Industries, Siemens Energy, Siemens Mobility, Siemens Healthineers, and a few supporting corporate departments cross-cutting.

Carsten is with Siemens Healthineers. I am with Siemens Advanta, part of Corporate Technology, which is cross-cutting across all the companies and aims to support all the companies and the business units within them.

We are a company with a history of 170 years now, and we have a history of very physical products. We have a lot of heavy metal, so to speak. So again, why do we care about frequent software updates?

It is not because it is cool or fancy, because there are the latest tools out there we like to play with, or because the Googles and Netflixes are doing it. It is because the world is changing. Our world is changing. Our Siemens portfolio and our market are changing significantly.

This is quite a big change happening at the moment. There are two keywords here. One is digitalization, meaning the increase of the relevance of software in our portfolio and our products. We have new cloud-based offerings and products popping up in all our business units. We have mobile apps that become a critical part of many applications. We have a lot of on-premise software applications, and we have a lot of embedded software in most of our physical systems. The whole arena of software is constantly increasing. It has been increasing for many years, but it is increasing at a constant pace and becoming more and more important, meaning more features and more customer value are realized in software.

The second keyword is connectivity. If we run a cloud solution ourselves, we by default have connectivity to this product. But also for our on-premise software and physical embedded products, we have a tremendous amount of connectivity already today, and this is constantly increasing. If you add big data analytics, AI, deep learning, digital twin, and many modern digital technologies, you realize that our whole portfolio in many areas is shifting in a new direction.

If you look at this more closely, you realize that frequent software updates are possible for many of these scenarios. Moreover, they are not only becoming possible, they are becoming mission-critical. They are becoming really relevant for us to stay competitive in these markets.

Let me give you some examples of typical application scenarios within our industry. Starting on the left, we have a lot of new cloud-based offerings popping up in all our business units. If you run a cloud-based solution, customers today expect that you constantly improve it, and that you do not only run a yearly update or release as maybe in the past on more classical products.

We also have a lot of data-driven, value-added services, remote services based on data analytics that provide additional value to the customer. In this case, we could try to build the perfect product, spending two or three years developing it. The much better approach is to start with a minimal marketable product the customers can already use, and then step by step, on a continuous basis, increase the functionality, make it a better product, add more modules, add more value to the customer, and thereby increase the user experience based on frequent, regular enhancements.

We also have on-premise devices. If we have the capability for frequent releases, we are in a position to do targeted advancements of these devices, either on a regular basis or on a demand basis, depending on the scenario.

There are also big data analytics platforms, where we have a lot of business applications on top coming from the vast amount of different businesses we have across the different companies within Siemens. If those applications want to evolve and have additional demands on the data analytics platform, we need to be in a position to evolve the data analytics platform at a high pace with very frequent updates in order to meet the demands of the business applications depending on it.

AI-based systems are also increasingly used within our products. We all know that AI algorithms are innovating very fast these days. In order to benefit from those fast advancements, we need a continuous exchange between implementing new and improved algorithms in production, learning from them, feeding the learnings back to development, and then continuously improving. If we only did this on a yearly cycle, our learning curve would be very flat. If we are in a position to do it on a biweekly basis, for example, we have a much steeper learning curve, which is crucial to utilize these new technologies for our benefit and for our customers' benefit.

And we have totally new digital businesses where we want to use experiments in production to explore real customer needs, in order to give us the right direction to evolve these new businesses.

For all of these scenarios, although they are different in some sense, they all have in common that they benefit from continuous delivery and DevOps approaches because we need frequent updates. I would even go one step further: they really depend on it. We will not be successful in these markets if we are not in a position to use these approaches and technologies to increase the frequency of our updates.

Applying those approaches in a company like ours is quite a significant change. You can see this best if you look at where we are coming from. For the last 30 or 40 years, maybe even more, in an industrial company like Siemens, in a large complex organization, everything was optimized based on a project approach, whether a 6-, 12-, or 18-month project. This does not only affect R&D. It also affects all the downstream interfaces to operations, sales, service departments, and adjoining supporting functions and services like patent analysis teams, technical documentation, business administration, and so on.

If you really want to move from this project-based approach to continuous value streams, the key mantra is: stop thinking projects, start thinking continuous value streams. It has a huge, significant impact on many aspects that should not be underestimated in our case.

Therefore, we created a picture that we typically use to trigger awareness for the scale of this change. On the left-hand side, as you could see from the different scenarios I described, there are different application scenarios for DevOps and continuous delivery approaches within our businesses. They are all moving in the same direction: more frequent updates, feedback loops, and a higher degree of automation. But the concrete implementation is different.

The first step is to understand where we want to go in a certain business and portfolio area, depending on whether it is regulated or non-regulated, on-premise or cloud, and to become aware of the desired target state we want to reach in three to five years. Then we have to look, starting from the business model: how do we interface with our partners? What are our customers' concerns on this topic? We need to work on the processes, especially release processes. We also have to talk about roles and responsibilities. We have new tasks coming from cloud operations. Who is doing what? How are our regions involved? How are we running field service in this context?

There is also the technological part, where we need to think about delivery pipelines, cloud technologies, and software architecture. For example, we can have stronger decoupling between safety-relevant and non-safety-relevant parts in our product. Then we could enhance the frequency of the non-safety-relevant parts at a much higher speed than the safety-relevant parts.

All of this has to be complemented by a transformation and transition program approach, change management, with a strong emphasis on addressing cultural changes. This is a big thing we are pushing here. I think it will keep us busy for the next couple of years. But we are strongly confident and convinced that this is the right direction we need to move in, because our world is changing; therefore, we need changing product development.

Carsten, can you give us some insights into how this transformation is managed at Healthineers?

Carsten Spies

Thank you, Peter. Yes, we in Siemens Healthineers also have a long history in product development. It all started back in the 1800s with the famous Mr. Röntgen and the first industrially manufactured X-ray unit, and has now grown into various businesses such as medical imaging with computer tomography, magnetic resonance tomography, X-ray and ultrasound, as well as laboratory diagnostics and point-of-care testing, healthcare IT, and solutions for digital services.

Meanwhile, we are a global company and a market leader in the majority of our businesses. We are more than 53,000 employees globally and have an installed base of around 600,000 medical devices. But the impact our teams generate is better described with another number: more than 240,000 patients all around the globe get in touch with our products.

These days, digital transformation is a big topic in healthcare and therefore also for our product development. One of the answers to the digital transformation challenge in healthcare is the teamplay digital health platform, a platform built to transform the way healthcare professionals get access to the latest and greatest software innovation.

Back in 2014, we started to develop the teamplay digital health platform, mainly to enable ourselves to efficiently build new digital services in the area of operations management, clinical decision support, and patient management.

Let me give you two concrete examples. One is teamplay Insights, a service in operations management that turns huge amounts of data in radiology into insights and provides interactive data analytics visualization. Another example is AI-Rad Companion, a clinical decision support solution that helps radiologists interpret medical images with the help of AI.

Over the last six years, we have grown the teamplay digital health platform into a comprehensive platform tailor-made for healthcare. Meanwhile, we even opened it to partners, and we are proud that other med tech companies and startups have decided to use teamplay to build and operate their solutions.

While developing teamplay, we actually started to transform the way we develop, deliver, and operate software. We switched from developing products into developing services. A software developer these days in our team has to really think about how to operate a service in production. We all learned together how to apply continuous monitoring to our services, and even to look at constant utilization of them.

We changed from long-running software development projects to short release cycles. In the past, a software development project within Siemens Healthineers took around one to two years. Now, with teamplay, we have a three-month mainline cycle to production.

To achieve this, we also shortened feedback loops and introduced new feedback loops into our development process. This starts with unit testing and goes all the way down to real customer feedback that we integrate into our processes.

Very important, and one of the key success factors of the whole journey, is to make teams of product managers, software developers, implementation engineers, and operations engineers work together. We can see this when we go together to customers to get their feedback. Together we do trade fairs. And I can really say that if I join the office in a day, I am proud that in teamplay we are really one team.

As we develop in the healthcare sector, we work in a highly regulated environment, and our processes need to comply with a variety of regulations. To be able to apply continuous delivery and DevOps in our domain, we were forced to tailor common practices to our needs. This is what we call continuous compliance. It is all about being able to deliver value to customers continuously, in high quality, and always comply with the regulations that apply to us. For example, we need comprehensive documentation of each and every individual engineering step throughout the process.

Let's look at customers and their benefits from new digital services and continuous delivery of value. Take IATROPOLIS, a hospital in Athens, Greece. The hospital is using teamplay applications in operations management to comply with changing regulations in healthcare, improve quality of care for us as patients, and improve patient throughput. With teamplay, we provide an easy-to-use service to them. With our ability of continuous delivery, we can incorporate customer feedback and changing regulations at a much higher speed than we are used to in traditional IT products.

With this, I would like to hand over back to Peter.

Dr. Peter Fassbinder

Thank you very much, Carsten, for providing us insight into a concrete example. I think it is very exciting that teamplay is really one of the most innovative product areas we are driving within Siemens, and I am also quite happy that I can contribute a little bit from my role on this one.

Gene asked me to also give you some insights into my role and how we are driving such a large topic as a tremendous change across Siemens. You could say we have established a couple of interconnected, interlinked circles. It is not an organization, it is not a hierarchical structure. It is more different networks we establish, or that establish themselves, and that are driving this topic forward.

Let me walk you through them bit by bit. We have established quite a few communities of practice, some within individual businesses and others company-wide. We also run large internal conferences on DevOps topics. We have many awareness events where we talk about this. For example, we established a regular webcast, especially in these times, where on a weekly basis we continuously deliver continuous delivery and DevOps knowledge to our community.

We also have a number of pilot projects within business units that are driving this topic forward. We have established a company-wide expert group for DevOps, with representatives from all the companies, that I am heading. We are also in regular exchange with external expert networks, for example at conferences or similar events.

We also have corporate key experts, corporate development, Siemens Advanta, and these departments where we are pushing things forward from a corporate side. This is also my role. I am a key expert for PLM process innovation, and in this context I am pushing forward the DevOps topics as part of my role.

The key idea behind these corporate supporting functions, corporate key experts, corporate consultancies, and departments is twofold. One is that we run concrete implementation improvement projects within individual businesses, while our expertise and best practices can be shared across the whole company. We really drive current issues forward and help improve the businesses today. But the idea of the corporate key experts is not only to think about improving today, but also paving the way for the future.

As a corporate key expert from a certain level, it is expected that you are also a thought leader in your field. That means you really look at what is behind it, what is coming up next, what new approaches and methodologies are emerging, where the industry is moving, and then pick up these things and push them across Siemens. I think this is a strong pillar of innovation capability within Siemens.

What is very important is the business relationship between these corporate functions, corporate key experts, and the business units. It is not that we drive topics just because they are our hobby and we like them. The businesses and business units have to pay for our services. It is like an internal consultancy, thereby ensuring that we really do things they gain value from. This model, which has been established for decades, ensures that we are not doing things for academic purposes, but doing things that are of value for the business unit.

Speaking about this corporate key expert role, I also have to admit that I really enjoy this cross-cutting role. I have been doing this for over 15 years as an expert driving things across Siemens. Every day it is still very exciting because I can deal with so many different teams, business units, and product scenarios, and I can create an impact all across Siemens. It is very rewarding every day I do this job.

Carsten, could you maybe give us some ideas on how you as one of the business units benefit from these networks and the corporate functions?

Carsten Spies

Of course, Peter. Within the business units, we benefit from the expert know-how of you and your colleagues, and of course the bandwidth and dedication you can spend together with us in supporting the transformation within our projects and products. Additionally, we very much like the best-practice sharing within the company, also in total as Siemens as a corporate. This is very helpful for us, and we use it broadly within our teams.

Dr. Peter Fassbinder

As a closing, we want to ask you: let's get in touch. Let's talk about continuous delivery and DevOps, and share our experience on how to apply it best in regulated industries. We are especially interested in your approaches to tackle continuous compliance, in the tools you are using every day, and in your ideas for how to tackle the cultural-change challenge that comes along with the transition toward continuous delivery and DevOps.

With this, we are looking forward to your emails or LinkedIn messages, and I would like to thank you for your attention to our talk. Thanks a lot.

Carsten Spies

Thank you also very much from my side, and I hope to talk to many of you soon.