Interview summary
The first use case is about Software AG’s collaboration with compressor manufacturer Denver Gardner. This company has 16 brands worldwide, for which Software AG supplies the system and its own mandates. Bernd Gross explains the business model they work with and how technology is used to achieve added value from which the end customer can also benefit.
The second use case is the example of the Swiss roller shutter manufacturer Griesser, which specializes in shades for industrial applications. Since severe weather has so far regularly resulted in high insurance losses, the company has expanded its business model to include a digital service. Whereas the core business initially consisted only of the manufacture and supply of roller shutters, these can now be raised on the basis of weather data, thereby preventing storm damage.
Finally, Bernd Gross reports on the use case of a copper cable manufacturer, which uses EDGE Realtime Analytics to detect quality problems faster, and reduce its waste. In this context, Madeleine Mickeleit and Bernd Gross, among others, make the connection to the topic of sustainability. This is an area where there are exciting solutions that can benefit enormously from IoT application and is also extremely relevant.
We will also touch on the topics of the three phases of IoT integration, how digital transformation can succeed from Bernd Gross’ point of view, and where Software AG Cloud solutions are used today as a white label.
Software AG is headquartered in Darmstadt, Germany, and is a global market leader for software and integration solutions and related services. It promotes the digital transformation of companies and enables rapid innovation in the Internet of Things so that customers can afford to perform agile pioneering work in differentiated business models. In return, the company provides the freedom to connect and integrate any technology from apps to EDGE. Software AG has nearly 5,000 employees and is active in 70 countries.
Podcast interview
Bernd: Our market version is a holistically networked world. So not just connectivity, but also in the B2B context, related to enterprise IT. This is our strength and where we already have an incredible number of customers. We are now combining this strength with our cloud offerings. This is our USP, which many customers appreciate. And an IoT project is IoT on one side, but integration on the other – integration into the processes, integration of the employees and of the new business models. This plays an important role, and this is where we are one of the world’s leading software providers, with the special combination of IoT and integration.
Madeleine: Now, in my podcast, I talk specifically about Industrial IoT. At the end of the day, it’s nothing other than a digital transformation that I’m undertaking with the company. What is your current assessment of the German market and the commercialization of these projects? What is the current status quo?
Bernd: Tesla is a good example of this. Because what Tesla has delivered is a customer experience that is improved every few weeks via software updates. I get a new software release every two weeks, new functionalities, new games, etc., always up-to-date. For example, since the other day I’ve been able to choose between the different locomotion options in my radar system, that was put in with the last update. That’s an exciting thing, considering that Tesla has a service-oriented business model and doesn’t see itself as a car manufacturer in the classic sense.
The Volkswagen ID.3, on the other hand, as the flagship of Volkswagen’s electromobility, is certainly also a super car, but it cannot provide software updates. To have new software installed, you have to go to the dealer’s garage.
You can see that something has to be done. A mid-market OEM, a machine builder, a compressor manufacturer (e.g. Gardner- Denver, one of our customers), they just need to think more software, value-driven over technology. That’s what’s driving the market right now.
Madeleine: In mechanical engineering, it is often the case that I also sell a digital service with my machine. However, I first have to develop it with my customer. Are there already working models on the market? What is our position?
Bernd: What we are talking about here is a digital transformation. Through IoT, digitization is brought from the back office to the front office, towards digitization, machines and the physical world. This is basically a new generation of digitization. The task now is to shape this transformation as a company in a positive way. One methodology that helps determine whether or not the project will be successful is co-innovation with the end customer. If I don’t include this from the beginning, I’m already off to a difficult start. But then agility is certainly also important to continuously improve. These are all important points, but the most important thing is to involve the end customer, otherwise you can’t transform as a business We sometimes tell our customers we will only co-innovate with them if they involve their end customer directly.
Madeleine: To understand your business a little bit further, do you have an example? Can you share a real-world use case to show what role you play as Software AG?
Bernd: One example from the medium-sized OEM environment is the compressor manufacturer, Gardner Denver, which has been a customer of ours for several years. It is connected to 20,000 compressors worldwide, and must create added value for the end customer through networking.
To do this, they started with simple monitoring services and gradually integrated other tools from us.
Gardner Denver operates globally with over 16 different brands for which they need to deliver digital service. As a native, multi-mandate platform, we certainly have the ability to deliver a mandate for each brand.
Ultimately, however, this is a closed system. We don’t need different systems per brand, you can offer everything with one instance and one system. It’s a step-by-step journey that you go through together. It starts with connectivity and continues with dashboarding and monitoring. Now we have just launched a predictive maintenance module, as an additional value-added service.
Madeleine: You just said that the bottom line for the customer is to create added value. What does a business model like that look like now, is it going in the direction of paper use?
Bernd: They have a freemium/premium model, which means connectivity is free and part of the offer. Other services, such as predictive maintenance, paper use or premium services are also marketed. This works via a monthly subscription fee, how high it is varies. In the factories in the compressor environment, there are sometimes huge rooms with lots of compressors. They can provide predictive maintenance for the entire system with valves and chillers, for the entire installation. That has significant added value, and that’s where they sell that separately.
Madeleine: Before that, someone must have handled these things manually. Now I have connectivity to my compressor and a new model that I offer. Who is the end user there? Is that a service employee, or what persona are you adding value for?
Bernd: Exactly, for the service technician and the service expert on site. With the trouble ticketing system, this is then handled via a notification. The first persona is really the one who does the service, the maintenance. But also the one who wants to have the production figures. For example, the production manager also has a dashboard. So over the years, a wide variety of personas develop around the digital services, services and applications around the compressor.
Madeleine: That’s actually the core added value of the IoT. Because maintaining a system via a ticket system or a VPN tunnel has also been done before. Now the added value would be to work with the data and create additional added value. As you said, there’s the production manager, for example, who sees additional insights through certain data correlations, and thus again has potential for cost savings.
Bernd: In many plants there is the situation that you have a spare device in case something fails. Technicians then connect the technology from the plant to another piece of equipment to reduce downtime. The better predictive maintenance now becomes, the less need there is for a spare part compressor. These are already considerable potential savings that can then also be offered to end customers.
Madeleine: Now you’ve said you’ve been on this journey with the client for two and a half years. After all, you don’t start directly with predictive maintenance; you first need data that you can work with. From your experience, is that the amount of time it also takes to develop something like that, and to integrate solutions like that?
Bernd: In principle, I see three typical phases in an Iot project. The first phase is the connection phase (“stand-alone”), where I focus on a single application. Independent of the IT systems and other processes, I build a stand-alone solution. Simply to get a feel, implement the connection and manage the first application. That’s about the first six to nine months. When I set that up, I also have some sense of the complexity. Because the irony is that an IoT project initially increases complexity due to the new data networking. Because I have to make sure that this is continuously available. I need to be able to rely on the new data before I start to optimize the processes.
This is the second phase, which also lasts another six months. There, the data is integrated into the existing IT landscape and processes are optimized. Efficiency and productivity is increased. This second typical phase can only be done when you really have “reliable data”. So SLA data, as a good and reliable basis. Only then can the process be optimized in a meaningful way.
The third phase then focuses on the topic of new business models and innovations. At Denver Gardener, we had the innovation Predictive Maintenance, of the entire compressor room. With many, we also have new business models, things like asset-as-a-service. They are also a factor. So, in summary, three phases: Stand-alone, integration into existing processes, and then new innovations.
Madeleine: What’s your take on existing infrastructure? For example, for customers who are in the process of implementing an MES system. Would you start an IoT project in parallel? Do you wait until the existing infrastructure is in place and then add an IoT solution on top? How do I go about it?
Bernd: With Industrial IoT, you have the traditional pyramid of automation. The foundation is the sensor technology, then comes PLC. Then there is a Scada monitoring area, then comes the MES system, then at the top is the ERP system.
With the IoT, this pyramid is changed. In IoT, a Distributed Architecture with Cloud and EDGE is often used instead of the pyramid. This combination of cloud and EDGE then often attached to an MES system. You can then already run new applications or analytics on site, with an open microservice concept. So a container architecture on top of EDGE and the cloud. The traditional automation projects in IoT have already been implemented by many. Now these architectures are being added in parallel. Like a Distributed Architecture, where I have an open EDGE platform and can innovate on the spot.
I recommend a strong focus on new technologies. Because it simply makes more sense, there are many innovations and investments. For example, many self-service tools. In the old world, you have to virtually develop everything you have in the user interface yourself. In the new world, there are applications where you can easily create workflows and integrations via self-services. You can even trigger your billing systems. Possibilities that are immediately available to you in the new world, would have to be developed expensively in the old world.
Madeleine: Yes, absolutely. Now you have given an example from the compressor environment. Do you have another example that you can give us to go back to a little bit on other business models as well?
Bernd: We have over a thousand IoT end customers who use our technology worldwide. For example, Griesser from Switzerland, which focuses on shading in the industrial sector. They started with a simple use case, a simple smartphone app for electric shutters. Then they just kept going and kept bringing in more innovation.
For example, there are an incredible number of insurance claims during severe weather. Now the shades are automatically raised when a storm with hail or high winds is approaching. This has enabled them to achieve significant savings on insurance claims. These are relatively simple use cases to me, but there are thousands of them that we’ll see in the future.
Madeleine: Supplying the shutters was effectively the core business before ? Simply as a classic on-site installation, and now basically with a sensor that transmits to the cloud? Through mobile connection and that correlates with data on site, and then I can use a corresponding service on Cumulocity that tells me how weather data and external data correlate, or how does that work?
Bernd: Yes, that’s right, it’s mobile-connected. That’s then integrated into the smartphone app, and you don’t even know that you’re using Cumulocity in the background. The user interface has everything in it, and then you can optimize automatically, that’s the idea. So that you simply do not have problems with weather etc. in the future. There are, of course, many more ideas that will play a role in the future. But this is an exciting field, where you can correlate with external data, so to speak, and thereby get your added value.
People often think only of machine data in the industrial IoT space, but in reality you have to think much bigger. Customer data, supplier data, supply chains, a whole host of external data that I can now combine easily and cost-effectively. It’s kind of an investment, because once I’ve gone down a path like that I just have APIs that I can connect and put logic on there.
For example, we offer real-time streaming/streaming analytics. This is a modular tool that you can easily deploy via a micro-service with one click. Then you can model your business rules yourself through a user interface, and then you deploy those rules. Nowadays, the step to make this investment is really minimal. By doing this, we enable a lot of innovation to go in that direction through these little barriers.
Madeleine: And above all, networking things, like the sensor in this case. This is also the smallest sensor technology, where I then also have the option of networking the OT world with the cloud via certain services.
Bernd: I know you specialize in the Industrial IoT space, so I have a few examples for you there. But I see really exciting use cases that can be abstracted into the Industrial IoT world. We have many telecommunications customers who use the whitelabel version of our website, and brand it themselves. Here in Germany it’s the “Cloud of things,” in Holland it’s KPN, in Austria it’s A1, in Australia it’s Telstra, or NTT in Japan. There are really a lot of differences there, they then in turn have dozens of end customers using that platform.
There is then a very simple example from KPN concerning bicycle insurance. This works great in Holland, bicycle theft is a big issue there. They have a retrofit kit, you stick that to your bike with industrial glue and that’s transmitted over narrowband and that drops your bike insurance by 50%. They have a service that someone will search the bike for it, and for that the insurance company offers half the insurance premium. Such simple use cases are stimulating; you can think about them further and abstract them. How might I use these in other areas? That is very exciting at the moment. The growth of the market is enormous with over 50% in the last quarter. That’s a great thing, of course, but it also reflects the market. We are really in commercial deployment. We’ve been dealing with pilots and bots for many years, but I don’t see that happening anymore.
Madeleine: This is the closed loop that i was referring to. The commercialization curve is now slowly going up. I’m curious to see what exciting topics will emerge in the future. Now you have already seen various projects. Do you have some tips and tricks from practice where you say mistakes could definitely have been avoided and costs saved?
Bernd: Yes, definitely. The phases we have just discussed are very important. Ultimately, everyone has to decide for themselves how many phases to divide this into, but you should think step-by-step. The biggest mistakes I see are when companies start using AI algorithms now, they have now searched for data, set up cloud warehousing and are now building fancy models. Personally, I always have a stomach ache there. Because I know people just underestimate this new technology, IoT, the management of these connections, so that everything runs cleanly in the end. What good is the nicest algorithm if I don’t get the data in the quality I need to use it at all? I see big problems in starting off too ambitiously with futuristic ideas.
The second thing I often see is the business cases. If I assume that a journey will take eighteen to twenty-four months, then I should communicate that as stakeholder management. The person who is responsible for the IoT project and is now going from an offline to an online business model should make sure that they communicate that solidly, that this multi-year journey is not going to happen overnight. The digitalization of the new generation, the holistically networked world. I very often see that it’s not the projects that fail, but the business models, because people were too overambitious. For me, these are two lessons learned. On the one hand, too ambitious on the solution side, with artificial intelligence, etc., and on the other hand, too ambitious in the business model.
That’s where I think you find a partner that has a PSE-use business model. Then you can also offer this out front, and grow over time. Via the connection, but also via the use cases and applications. But you should know that this takes time. While it’s nice to have recurring revenue coming via a PSE use, recurring revenue is harder to work up than onetimers. Communicating that correctly is not always easy. Mechanical engineers in particular always think in multimillions or billions. If they see relatively small numbers with recurring revenue, then they are missing the vision. The fact that there will be exponential growth at some point, and that recurring software sales have a completely different profitability than, for example, one-time machine sales, are difficult discussions where companies struggle.
Madeleine: Yes, absolutely. I also briefly touched on incentivizing sales in one of my last sessions. Of course, this is then also a completely different objective. Do you think it makes sense to detach something like that from the core business? These are, of course, major strategic decisions on the part of the management. But it’s often two worlds, from offline to online, do you think it makes sense to separate that from the core business? To build a separate company, and build the internal structures on top of it?
Bernd: That’s the question, you probably can’t answer that generically. I think this new digitization demands a CEO buy-in. This is a boss issue, something no project team should be doing. This means, that it is actually about the survival of the company. If I don’t adapt and create the new structures, I will most likely have a problem with my core business in the future. You can no longer think offline and sell anything without managing the experience. You have to rethink and reset that, which makes it a boss issue for me. This has to be present and incentivized in the board, in the CEO. For many companies, that means it’s part of the core, and that’s where many struggle. The new structures do not fit into the old ones. Therefore, there must be an executive buy-in.
This is needed so that you can successfully implement the project. How you solve this depends a bit on the industry, there are different approaches. In the retail sector, this is solved within the organization; in mechanical engineering, there are separate software units. Volkswagen, for example, has seventy-seven different software teams, and they have now merged them. What I think is good about Volkswagen is that, for example, Herbert Diess, the CEO of Volkswagen, is fully behind it. That’s the boss’ issue, he’s transforming the company. That is the primary deciding factor for me. How I then ultimately finalize that is secondary.
Madeleine: It’s insanely exciting, these are probably industry topics now where that works. That probably also depends on the expertise, what the company wants to build up in-house, or what you do externally, depending on what is coming in the future. Do you see any new trends that will emerge if I now focus on the next five years?
Bernd: The market version we have is this holistically networked world. This is a version we can use to address megatrends. For example, climate change, quite separate from technology, that’s a huge issue. Probably the biggest issue we need to solve in society. That’s gone down a bit now because of the pandemic, but it’s certainly coming back. Every company must ask itself what it can do and how it can reduce its own carbon footprint. We at Software AG are also active in this area. We use technology to reduce our own carbon footprint. But of course, we also offer our digital platforms cumolocity.io, etc. so the integration platform and the IoT platform, which are merging together at our company to help other companies reduce their carbon footprint as well.
We are doing this, for example, with Smart City in Dubai and with Smart Water Management in Australia. Water is a critical resource in Australia. In the old world, you waited until someone called and said, “there’s quite a bit of water running across the street here, something’s broken.” In today’s world, what you see is fully automated water distribution. So if there’s a burst water pipe somewhere, you’re immediately alerted to it, you immediately know where that is, and you can send your teams there. You immediately see water depletions, abnormalities, and that kind of thing happening more and more now. As a software provider, we also want to play a role in this and supply our technology to make a positive contribution.
Madeleine: That’s also part of the beauty, that you can also create great sustainable solutions with this technology. We have now set up a separate category on the subject of sustainability, because we now have such great examples of this, including from our SMEs.
Bernd: We have two examples from our customers in the industrial sector. One, is a copper cable manufacturer, and they can immediately detect if there are quality problems somewhere, and stop production, through these EDGE real-time analytics that we have installed there. Previously, production continued for 1-2 minutes and several more kilometers of copper were produced, although the copper lines were no longer good quality. The company then had to cut that out as waste. Now they can analyze and stop production immediately in the range of 100ths of a millisecond, resulting in only a few meters of waste. That’s 99% less emissions. When you consider that, that’s of course enormous.
Another example is DÜRR, one of the world’s leading automotive suppliers for paint plants. They have analytics, and can tell via the robots’ data analysis whether the quality of the paint job is flawless or not. And this is done at a high frequency, 100,000 data points per second are processed in our EDGE technology. That’s better than the visual inspections employees used to do, on every 5th or 10th vehicle body. They can now automate all that and recognize flaws immediately without anyone having to look at the vehicle bodies. That’s big savings and a huge relief.
