Investing in renewable energies and infrastructure – but intelligently!

A warm hello to the first episode after the summer break. Today we are looking at a new and exciting topic in the field of energy management. It is all about an innovative method of evaluating and rethinking data, especially when investing in renewable energies. This includes not only IoT data, but also information from areas such as purchasing, controlling and production. The aim is to optimize investment decisions and rethink the energy infrastructure.
Today I’m talking to Joshua Kuepper, co-founder of RIZM. He brought along two interesting customer cases from Schaeffler and BMW that show how their solutions are used in practice. Perhaps you should also rethink your energy infrastructure after this episode. I hope you have fun and look forward to exciting insights and concrete use cases. As always, you can find all the information at iotusecase.com. And with that, let’s get started with this episode!
Hello Joshua, welcome to the IoT Use Case Podcast. Who are you and what exactly do you do?

Joshua

Thank you very much for the invitation. I am very happy to be here. I’ve been following the IoT Use Case Podcast for a long time and I’m happy to now have the opportunity to share what I do and who I am. I am Joshua Kuepper, my background is in actuarial science and risk management and I am one of the founders of RIZM. We support industry in the energy transformation and at the same time in maintaining its competitiveness. Specifically, I am responsible for the market launch of our technology. We have two other founders, Elias and Philipp, who have a research focus and have been researching how to optimize energy decisions with data. My job is to ensure that our solutions are successfully implemented for our customers.

Very nice. Greetings to everyone at RWTH Aachen University, to the founding team and to Düsseldorf. You’re based in Düsseldorf, right?

Joshua

Exactly, we are represented at several locations. The main office is in Münster, but I have set up an enclave in Düsseldorf. This works well because the locations are not far apart.

Are your customers mainly from Germany, or do you also have international companies?

Joshua

We work globally, but the majority of our customers are headquartered in Germany. Many of them are DAX 40 companies that also have their headquarters in Germany.

Maybe you can give us an overview of the projects we are talking about today, or share a few success stories and use cases where companies are using your solution.

Joshua

Gladly. I start a little more generally and then get more specific. There are two central starting points: Firstly, if renewable energy is produced, this energy is unbeatably cheap, provided it is produced. This is the case globally, not just in Germany. This is leading to an ever-increasing share of renewable energies in the energy systems and thus to greater price volatility, which we can also observe in Germany.

A quick interim question: by energy systems, do you mean the systems in use in industry?

Joshua

Also, but I mean, for example, the European electricity grid or the electricity grid in general. This price volatility will increase, that is a trend. In some cases, we see absurdities such as negative prices, which we observe in Germany and other countries. Secondly, there are strong commitments from companies to produce in a climate-neutral way. This means that they have to produce cost-efficiently and keep the costs per unit as low as possible. The solution is to maximize energy consumption when energy is cheap and green and minimize it when it is expensive and less green. This sounds simple, but it is a major challenge, as production should not be restricted. This leads to many questions, such as the optimal design of the energy infrastructure and the selection of the right procurement channels. The flexibility of the infrastructure also plays a role. We will explore these topics in greater depth using specific use cases.

Exactly, we’ll talk in a moment about how the infrastructure can be optimally designed, as well as about procurement and flexibility. You have already raised some great questions, which we will address in detail. But sorry, continue with the examples first. I think you wanted to name a few more.

Joshua

Exactly, I’ll just give you a few examples of the direction in which the resulting questions are heading. A good example is what we have implemented together with Schaeffler. Our software was used by creating a digital twin of the relevant energy system. It was discovered that improved production scheduling in combination with active load management enables Schaeffler to produce products at certain locations when energy is cheap and available. Of course, there are limitations, such as production times and full capacity utilization, but the software was able to identify exactly the cases where it works. This enabled us to realize a business case with savings of around 50,000 euros per month at one location. This shows that money can be saved without CapEx investments simply by linking the energy market and production flexibility.
Another example goes in a completely different direction, namely infrastructure. What we have achieved together with the BMW Group is also a success story that can be requested on our website. BMW had a high CapEx requirement to achieve the targets they promised their customers. However, our software has shown that some of these investments are not absolutely necessary and that there are more optimal energy systems than originally assumed. The result was that BMW was able to save a very high three-digit million amount. Some of this relates to CapEx, some to OpEx, as investments can be made differently and the energy system becomes cheaper while still achieving the targets. This is important because many companies reach a point where it becomes financially difficult to meet their obligations. If you can save a high three-digit million amount, it will be much easier to meet the obligations.

Thank you for sharing these two concrete examples. Using the cases of Schaeffler and BMW, we can now delve deeper and better understand the challenges and business cases behind them. Before we start, I wanted to ask one more question: What role do you play as RIZM, and what do companies like Schaeffler or BMW do themselves? You mentioned infrastructure and so on.
What exactly do you provide and what do your customers do themselves?

Joshua

The software we supply has two components. On the one hand, there is a business case calculator that not only allows you to find out the business case for a particular measure, but also which measures are the best from the many options available. This is the one thing we provide our customers with so that they can put their ideas and topics through a filter and ultimately make the most important decisions that they should take a closer look at for 2024 and 2025.

In other words, you offer a software-as-a-service solution that is integrated into existing infrastructures. You supply the software, part of which is this business case calculator.

Joshua

Exactly. The second part of the software makes it possible to implement the identified business cases. For example, if components have to be operated differently, such as an electrode boiler that has to be operated in a volatile manner. This means that the software includes both the identification and the realization of the cases. We provide our customers with intensive support at the start of the project by setting up the system together. Our Success team works with customers to utilize existing data, cleanse it where necessary and integrate it into the digital twin. We help to identify the quick wins and show that the whole thing works. Over time, customers become more and more independent and can increasingly use the software on their own. At BMW, for example, after two years of cooperation, they have set up their own department for planning energy systems and are now also establishing one for operating the systems. From this point onwards, we only provide support, as the customer uses the software independently. It is much more efficient for the customer to do things themselves and answer questions quickly than to always have to rely on external support.

Nice. Before we go into the business case and the typical challenges of your customers, one last question: You are here on behalf of the founding team. You’ve already mentioned it briefly, but what motivated you to found RIZM as an independent company? There are many solutions on the market. What makes RIZM special and what was your motivation for founding it?

Joshua

We have noticed that decisions are often not made algorithmically with the necessary complexity. There are many reasons for this. On the one hand, it is often too time-consuming to create a complex mathematical model for every question, answer the question and then implement it. This is sometimes very demanding. On the other hand, the integration of data complicates this process. Last but not least, the algorithms play a decisive role, as conventional planning often leads to higher costs and does not find the optimum operating point for the energy infrastructure.
Our aim is to support industry in becoming climate-neutral while maintaining its competitiveness. We wanted to create a solution that would enable the industry to answer important questions quickly with minimal effort and thereby remain more competitive and meet its climate commitments. That was the decisive reason for founding RIZM.

Very nice. You’ve just given me the perfect cue for the next question. You talked about costs and mentioned some big figures, for example at BMW and Schaeffler. Can you explain what the classic business case and business challenge is for many of your customers? Why is this important, and why are decisions not always made as they should be?

Joshua

Gladly. Let’s talk about our typical customers first, because that’s where the typical questions and business cases come from. We have a strong presence in the automotive sector and work with almost all major OEMs, as well as with many suppliers. However, we are also active in over a third of the DAX 40 companies, in some cases in all plants across the Group. Our customers also come from the chemical and steel industries, so we are industry-agnostic.
A typical challenge is that many companies struggle to achieve their climate targets because they find that the necessary CapEx investments and additional costs, CapEx or OpEx, reach a critical level. A key question is therefore whether there are more cost-effective ways of achieving these goals. The answer is almost always yes. We have shown that BMW and many other companies can do it much more cheaply.
Another issue concerns smaller, short-term measures. We can improve cash flow here by starting with small use cases, such as the use of flexibility or the optimized operation of systems. We then show the exact business case and combine it with other measures to create synergies, for example by combining purchasing and production, as in the Schaeffler example.

So one example would be to operate systems such as CHP units or other energy systems differently to the way they have been operated in the past?

Joshua

Exactly, a simple example would be to replace a gas boiler with an electrode boiler, a hot water tank and a heat pump. This creates flexibility, as the electrode boiler or heat pump can be operated as required. You can store cheap energy and use it later when it becomes expensive to cover heating requirements. With additional flexibility, such as ventilation or machines that do not run at full load, the case becomes more complex, but also more rewarding.
You can then quickly find out whether it is worth using these flexibilities, integrating PV systems or purchasing more volatile energy products. We often start with small business cases that can be realized without investment and use the savings achieved to tackle larger issues, such as whether certain infrastructure investments are still necessary. This makes the whole thing affordable.
The difference to other approaches lies in the way our optimization algorithm works. In contrast to purely simulative approaches, where you ask “What if?”, we answer the question “What should I do to achieve the best result?” We take a lot of data into account to find the best solution, whether it’s short-term operational optimization or long-term planning.

When you talk about data, I think of the volatility of energy prices and information that is available in certain databases. We also talked about larger systems. What data types do you typically need to implement your solutions? And do your customers normally already have access to this data, as many already use energy monitoring? Can you say anything more about this?

Joshua

Sure. Let’s start with why we need the data in the first place and which data exactly. We need the data to find out where energy will be needed in the future and where there is flexibility in this energy consumption – i.e. whether consumption is necessary at a certain time or whether there are time windows in which it can be postponed. The second question is where the energy can come from.
This is the basic framework. However, it can become more complex. We can use historical data to simulate what it would be like if we were to continue the past on a one-to-one basis. We can apply trends to historical data to map future developments, or use synthetic data if we know that something will change in the future, such as the production of a new vehicle or a different purchasing strategy in a few years’ time.

So synthetic data would be, for example, simulation data or data based on assumptions?

Joshua

Exactly. This could mean, for example, that I know the energy requirements for the production of a new vehicle or assume how certain components will consume energy in the future. In the chemical industry, this could be a column where I change the chemical process and therefore have a different requirement than historically. The more accurately I can map the future, the better the result. But I can also start with historical data as an approximation to create a basis.
Many companies already have energy management data, either at plant, hall or factory level, which is often sufficient for the first steps. This data helps to understand what the historical energy consumption was and how the demand could be optimally covered. It is important to emphasize that we are not an energy management system. At Schaeffler, for example, we work closely with autinityE3, a system that I can highly recommend.

I’ll link this in the show notes. We have already covered Schaeffler in one episode, and there will be another one this year.

Joshua

Exactly, what they are doing is really exciting. We use various energy management systems, but autinityE3 works particularly well. Basically, we use energy management data that is already available, as well as procurement data – i.e. information on existing energy contracts and infrastructure data, such as existing systems and possible expansions. This data helps us to create a digital twin that maps consumers, lines and producers.
We don’t have to work with live data, but can calculate business cases based on this existing data to find out which options are particularly attractive and which should be looked at more closely. Other important information includes resilience – i.e. how available the energy must be – as well as requirements for realization and amortization periods. Assumptions for controlling or the available CapEx must also be taken into account.
Finally, we need data on energy purchasing in order to decide whether an efficiency measure should be calculated with volatile or constant energy prices. We take into account the energy infrastructure, production requirements and any degrees of freedom. The majority of this data is usually already available. What we also provide are cost alternatives for technologies, parameters for alternative technologies and detailed modeling for plants and their operation.
We also integrate weather data, cost forecasts and purchasing forecasts, either provided by the customer or by us. This allows us to quickly compile business cases and analyze which measures are worthwhile. These can then be realized and connected in the same digital twin, e.g. via autinityE3. The digital twin can then make decisions about energy-related production scheduling as well as generate reports, such as the emissions per unit produced, which can be reported back to the ERP system.

I see. You therefore use data from different functional areas to cover all possibilities and implement different use cases. I understood that. Do you have a specific IoT use case where IoT data is frequently used? Is there a top use case that many companies implement, or do you tend to start with historical data and then build on this? Do you have a case that you often carry out with customers?

Joshua

A common approach we take with customers is to use existing data from the energy management system and the ERP system and combine this with purchasing data. Although purchasing data is available, aspects such as future energy prices when using spot markets, intraday markets or balancing energy markets are often not available in this form. We combine this information and analyze the potential in the existing energy infrastructure. We usually start with historical data, as this is easier than immediately building an interface to energy management systems. We export the data, create a simple digital twin and see how big the business case is and what makes sense. We then go deeper and connect the ERP system, energy management system and forecasts to automate decisions.
It is exciting to see how these twins grow over time. You start with a small use case that pays for itself quickly and contributes to your goals. Later, detailed efficiencies, e.g. of CHP units, are modeled so that precise data is already available for future decisions and well-founded decisions can be made.

Looking at the time, I have many more questions, but I would like to ask one or two final questions about your solution and the implementation. We learned how you go about it, what challenges you face and how your Software-as-a-Service solution helps companies to identify and implement use cases. Can you explain in more detail how your algorithms are used to better plan energy requirements with renewable energies, for example? How does this work analytically? That would interest me, since that is your core area.

Joshua

Our software builds a model from the existing requirements. The requirements are, for example, which energy must be provided and which flexibilities exist. You can imagine it like a room full of dots, with a line representing the requirements. These points represent all possible combinations of purchasing, infrastructure, production, storage, etc. The first line describes the energy demand that needs to be covered, taking into account the available flexibilities. The second line represents all procurement possibilities and restricts the range of options. Then there is the infrastructure – both existing and potential. The software solves the energy balance for each node in the system to ensure that the required energy is provided at all times. This applies to all forms of energy, whether electricity, heat, chemicals or compressed air.
The software takes into account weather availability and uncertainties through stochastics to find the optimal setup. It can then suggest investing in a solar system in two years’ time, for example, or opting for other technologies. The software shows the most cost-optimal point, i.e. where it is most worthwhile investing in infrastructure or using existing flexibilities. You can also look at shorter-term scenarios and find out what is possible without investment.
It is important to understand that it is no longer just about efficiency. It is okay to consume more energy if it is cheap and climate-neutral. We can therefore design systems in such a way that they utilize volatility and act in a grid-friendly manner. In addition, other constraints such as emissions can be included to ensure that the cost-optimal solution also meets the emission targets. The software then simulates the optimum solution and shows how it will behave in operation. This makes it possible to check whether the digital twin correctly reflects reality and whether the decisions can be transferred to future scenarios.

Okay, I sketched the picture you just drew on a small piece of paper. Do you have a link or a document that we can put in the show notes? I would like to link this so that our listeners can read it. And of course a link to your software. If you’re listening now and think these are exactly the questions we have, or you’re working on similar topics, just get in touch with Joshua. I think you have something we can link to in the show notes, don’t you?

Joshua

Yes, I would put together a small document that explains the key points – without any advertising. It contains a basic list of requirements for energy decisions, explains how numerical optimization works in comparison to simulation, and is presented in a simple and understandable way. I will gladly link this. I would also recommend anyone who is not yet familiar with it to visit the energy charts at energy-charts.info. There you can better understand the energy world and see why we have to make decisions differently today than in the past. We can also link this in the show notes. And if anyone is active in the energy sector and would like to talk about it, we are happy to share best practices without obligation. We have gained a lot of experience with various companies and are happy to pass this on without directly demanding anything major.

Perfect. Just look under Joshua Kuepper on LinkedIn, you can find him there. I’ll also put the link in the show notes. It is really exciting to see what RIZM is doing and will do in the future. This brings me to the last question: Where do you see technological developments in the field of IoT live data, especially in the energy supply, in the next few years? What is decisive here?

Joshua

We can see that the energy markets are becoming increasingly fragmented. It used to be enough to make decisions on a monthly basis, then it went to a weekly basis and so on. Today we are approaching a decision basis of quarter hours. This requires a high degree of automation. If I have to decide in less than a quarter of an hour whether a storage system is running or how my energy system is operated, everything has to be very well automated in order to exploit the potential. Volatility will not disappear – it is an opportunity that must be seized, which can be very attractive financially, but requires sophisticated automation. Many systems that can be operated flexibly must be controlled in such a way that they interact optimally with the markets. There is not only the classic spot market, but also balancing energy markets and many other mechanisms that need to be taken into account. The trend is for purchasing, production and energy infrastructure to be very well networked and work synchronously at less than quarter-hourly intervals. This is a trend that we see with our customers and that we are continuing to drive forward.

Very strong. I am looking forward to a follow-up to this episode because I find the topic very exciting. I still have a lot of questions in my head, but I wanted to give you and RIZM the stage first to show what you do. You have impressive customers and huge untapped potential. Thank you very much for joining us today. It was very informative and I think the concrete examples gave us a good understanding of where the challenges lie and how you tackle them. Network with Joshua and plan a follow-up meeting. Thank you for being there and sharing your knowledge. I look forward to a follow-up. And that’s the last word to you for today. Many thanks from my side for being there.

Joshua

Thank you very much for letting me be part of it. I am also looking forward to the follow-up. Thanks for the opportunity, and I look forward to hearing from everyone.

Very nice. Take care and have a nice rest of the week. Ciao.

Joshua

Thank you, ciao.