Energy storage at scale: C&I batteries and digital control

Stephan (00:02)
We have a lot of power electronics, but at the same time, we have less and less inertia, rotational inertia of gas turbines and steam turbines in the grid, which makes things more complicated control. And the digital technologies is not an advantage to have them, it’s actually the enabler for the energy transition.

Adam Sroka (00:40)
Hello and welcome back to the Hypercube podcast. Today I’m joined by Stefan Schnitz, who is CTO of RCT Pro. Stefan, great to have you here. be great for the audience just to understand a little bit about you, RCT Pro, what you’re up to, maybe a business they’ve not heard of and yeah, why we’re talking today.

Stephan (01:00)
Well, Adam, first of all, thank you for inviting me to the Hypercube Energy podcast. That’s quite an honor, in particular when looking at your previous guests here at the podcast. So I feel humbled. Thank you for that. Yeah. So about my background, I’m a physicist by training. I studied in Heidelberg in Germany, and initially I focused on theoretical physics, actually on theoretical particle physics, quantum electrodynamics for those who know a little bit about that.

So it’s quite theoretical, quite fundamental. And at some point it started to think and dream of arrivals and numbers and computer code. And I realized I need to change something. So for my PhD, I moved from Germany to Switzerland, to Zurich. So that’s where I still live and actually where we record, where I record right now to do a PhD in experimental solid state physics at ETH Zurich.

Adam Sroka (01:41)
you

Stephan (01:55)
I worked with a new material, so newly discovered back then called Graphene, which was actually discovered at the University of Manchester. So not too far away from where you are now. Yes, exactly. And basically we wanted to investigate this material for using it in future quantum technologies, for example, like quantum computing. That’s the buzzword today. So that was my PhD. Afterwards I decided to leave academia.

Adam Sroka (02:00)
Yeah.

Stephan (02:22)
For a couple of reasons, I joined ABB, Swiss Engineering Company, electrical engineering company, and I worked there as a scientist for close to eight years, mainly in the field of renewable energies then. So I switched from quantum technologies to renewables, and that’s a little bit what I did several times in my life, as you will see. And that’s also where I started working actually in the area of battery energy storage systems. So that was towards, well, maybe the last two or three years at ABB.

I left ABB to join a famous Chinese company, Huawei, here in Zurich. They started a new research center. So I was one of the first employees of Huawei. I helped building it up from zero to something like 100 employees during the COVID pandemic, actually. So they hired me actually for my background in quantum electronics. But because of my background or my expertise with ABB and the energy storage sector, I also was involved

several projects around power electronics and energy storage. Many people actually don’t know that Huawei is the biggest manufacturer of solar inverters in the world. So that’s quite impressive. And I left Huawei March this year and joined the company RCT Power. You mentioned RCT Pro, so there’s a difficulty I will explain to you. So I joined RCT Power in July this year.

So four and half or five month ago asked the CTO. So RCT Power is a company which is headquartered in Konstanz, Southwestern Germany, right next to the Lake Konstanz, but also next to the Swiss border. And RCT Power is a company which was founded 10 years ago and its first product was a hybrid inverter for the residential battery energy storage sector. So it was a very innovative product.

back then, but I think we will talk about it later on maybe. So we tapped into the C&I business or commercial industrial as well, and it’s totally different businesses. And we decided for several reasons. to spin out a new company that’s RCT Pro. It’s a working title that will probably be rebranding later on, but we just started this company beginning of this month, basically. So I’m one of the co-founders. I’m CTO of RCT Pro and we’re focused on the

commercial and industrial battery energy storage market that’s where I’m now.

Adam Sroka (04:41)
There’s plenty to talk about there. I significant danger. think I’ve got Yariv quantum electronics there. There’s a danger of this becoming a quantum electronics podcast. I have to ask are there any applications in battery storage for graphene?

Stephan (04:43)
Indeed.

Actually, yes, there are people discussing using graphene or multi-layer graphene. So the graphene itself is this single layer, this sheet of a single layer of carbon atoms. And because of the structure and the fact that it’s a purely two dimensional material, it has very specific properties, which makes it quite interesting for electronic applications. Like I mentioned quantum computing, but also for classical electronics potentially. When it comes to

to batteries when you think about electrodes, what you would like to have is a huge surface area. So you can take this two dimensional material and crumble together somehow, think, take it as an electrode. Typically today we use lithium ion batteries, we use graphite already, but it’s a standard graphite. So there are people playing around with graphene. To be honest, I don’t know what the is right now with using graphene, but it’s a super expensive material. So I don’t think when it comes to the…

Adam Sroka (05:49)
Yeah.

Stephan (05:52)
cost component, even if the performance would be better if it comes to the costs, I think it’s not competitive yet.

Adam Sroka (05:58)
I was at Manchester, I was studying at Manchester during the days of sellotape and pencil leads for for graphene discovery and stuff so was all the buzz there.

Stephan (06:04)
maybe

when were you at Manchester?

Adam Sroka (06:09)
I was there 2005 until about 2012. yeah, there was a lot of…

Stephan (06:14)
Interesting, because

we have met before Adam maybe, because I was at Manchester University in 2004 and 2005 as a RASMA student.

Adam Sroka (06:22)
Maybe just past just cross paths. Yeah, I funny that no, it was all the buzz then and then it kind of yeah, I’m not getting back at this but it’s Interesting, but then you’ve touched on as well. You’ve and had found a life right and yeah, this new business So I’d love to know yeah Well, I like I’d love for the audience to hear a little bit more about the journey from RCT power and what RCT power does and how it’s from inverters to other product lines and then

Stephan (06:24)
you

Adam Sroka (06:51)
kind of moving into RCT Pro on the next.

Stephan (06:55)
yeah, RCT Power, it’s a very interesting company. It’s 10 years old. It’s a rather small company. So their head count is, I don’t know exactly, maybe 60 people. So it’s a very small company. And as I indicated before, RCT Power pioneered to some extent the concept of the hybrid inverter. I don’t know whether they were the first company in Germany having a

commercial product, which is the hybrid inverter, but definitely one of the first ones. Now you have to think about 10 years ago in 2015, PV was quite common already, residential PV. Batteries were too expensive. However, a few people maybe had batteries, residential battery systems, right? But these were independent systems. You had the residential battery system and you had a residential PV system. Both PV and batteries are DC technology, right? So you have to have an inverter which converts a DC

electricity into AC, but independent systems and obviously from a technical perspective that doesn’t make sense. So you would like to combine the inverter to have one inverters to combine both systems. It would save money and you can directly charge the battery from the PV system. But that was not the state of the technology back then. So RCT power was one of the first companies to productize and commercialize the hybrid inverter where you have

two DC inputs or even more, but at least two, one for the PV system, one for the battery. And that make new functionalities possible. So that’s how the company started. That was the first product back then. It was around four to six kilowatts, which is pretty small for today’s implications. over time, we went to higher powers up to 10 kilowatts and their ongoing.

development works to push that even to even higher powers. Now we moved actually from silicon based transistors to silicon carbide based transistors in order to increase the efficiency. And then obviously on the digital side, we have an app, we have a cloud interface and all these kinds of things. that, but this was all in the residential sector. Now the hybrid inverter,

That’s 100 % made in Germany, which is quite unique actually today. So the team and Konstanz, which I led as a CTO until roughly September this year. So they designed the power electronics, they designed the mechanics, we do the programming of the software and the control, then the manufacturing is outsourced to contract manufacturer.

The battery itself is also our design to some extent, also the battery management system we program, but the manufacturing of the battery, what we call the battery stack there, that’s actually done by Chinese partner as for anyone basically today. Now it turned out that this company in China, our partners there, they launched in C&I business already some time ago. So for us, it became interesting to see whether we could also in Germany move into the C&I business.

and we partnered with this Chinese company on that. And that’s how RCT ProSummers started. we had, within RCT Power, we had this internal, let’s call it kind of startup, which tried to sell C&I storage systems in the German market. And this was actually quite successful. But as I said, it was a kind of startup atmosphere within RCT Power.

So we were trying things out, we’re moving fast, we adapted the system to customer needs, a very fast approach to things, to all these kinds of things. And it turned out that there were some differences in mentality of the people working in the residential sector, but also in the C&I sector, within RCT Power, there were strategic differences and there were some tensions. And at some point we decided, makes more sense to separate these two units into two different companies.

So that’s how RCT Pro started. And that’s where we are now. we founded the company officially about a month ago and we are in operation since November basically.

Adam Sroka (11:08)
What’s the emotional journey been like handing over CTO of RCT Power to someone else? Have you done that yet? Is that in transit? No. ⁓ Is it going to be able to? I think I’d find that really difficult to be like, okay, I’m concentrating on the next thing. Please don’t bring this. But I imagine you’ve got an incredible team and people you trust deeply.

Stephan (11:19)
No.

You know, the R &D team at the old RCT Power, many of the colleagues there have done this from the very start when RCT Power started in 2015. So it’s a very experienced team. They know the products by heart. so it’s an incredible team, very experienced.

I was hired as a CTO to basically bring in some new ideas from the outside, but also to really move the ⁓ expertise into the bigger storage systems. The experience I bring in was the one I had at ABB, for example, where I was the project manager for the new development of the utility scale battery system there, basically. So I was hired to bring in this kind of expertise and maybe some kind of

project management skills from large runnings, large projects. Now, since we separated these companies, the team at the old RCT Power can now focus on, or will focus on the residential sector again. So I think they have all the expertise they need. The guy who runs this department now, I know him, he is a very experienced guy. I think they will do a very good job. So I’m not worried about that.

For me personally, I mean, you were talking or asking about my emotions. I mean, the residential sector is a very interesting sector and it’s something where as the customers, they touch the system, they interact with the system. So I wouldn’t say it’s an emotional product and certainly not, but it’s a direct interaction between the customer and the product. And that’s nice, obviously, in the C&I and the utility scale, that’s a bit of a different story.

Adam Sroka (12:48)
you

Stephan (13:08)
We’re still in the same building. We talk to each other and we discuss things. So I mean, I’m still feeling not being involved, but still I know what’s going on. So that’s good. And on the other hand now with a new company, we will focus on C&I and later on utility scale. And my R &D team at this point is zero. But it means I can start and build up it from scratch. And that’s super exciting. I mean,

It’s zero, but we have some expertise. We are working together with Nextal IT company, which helps us with – and we’ll come back to that later on, I suppose – on the energy management system. I’m hiring an expert on the energy management system right now. So we’re starting.

We know the people we would like to work with, we’re in contact with them and their ongoing hiring processes. So it’s an amazing opportunity for us. It’s really fun.

Adam Sroka (14:03)
It’s an incredible market to be in just now. There’s so much dynamism in this. There’s things to do. There’s a lot going on. we develop sort of platform software solutions for like virtual power plants, all this stuff, optimization engines for home energy management systems and kind of residential scale. We do it for C&I and we do it for utility scale.

Stephan (14:16)
Yes.

Adam Sroka (14:26)
grid connected, big, batteries and things. And it honestly feels like three totally different worlds. they don’t from a, it’s funny because the foundational technology is the same, like it’s a battery, it just gets bigger. But the concerns, the way the products go to market, the way the revenue is generated, all these things like totally different. And the way the companies that we work with deal with them feel like aliens to each other.

Stephan (14:38)
Yeah.

Adam Sroka (14:53)
I just like what kind of opportunities and challenges have you kind of noticed that are there in that shift from sort of residential through to C&I?

Stephan (15:02)
That’s an interesting question. What kind of opportunities? Well, I think first of all, We see that the C&I market is growing quite significantly. So if you keep a small percentage within this market, but we keep it, we grow with a market that itself would be great. Residential market for a new startup company is much more complicated because it’s flat or maybe even shrinking, right? So in that sense, it’s simpler.

Now the opportunities is we know the local industry around Constance and also on the Swiss side and on the Austrian side. We know it pretty well for several reasons. There are private reasons. so we know the requirements from very specific small companies pretty well. And it’s this kind of knowledge which we would like to use, at least for some pilot projects to really

install products, learn the products, learn the customer needs and adapt to it. So for example, one of the customers we’re talking to right now, it’s a radiologist clinic. And so it’s something you have never thought about probably, but they have pretty high electricity consumption and they have peaks. And so this is like a perfect use case, for example. this is the kind of discussion we are having right now.

Adam Sroka (16:10)
Yeah.

Stephan (16:17)
⁓ and we would like to take out the installer in between because we think we can do it ourselves, we don’t need them, but then we can talk to the customers directly and understand their needs. Whether this will scale in the future, I don’t know. We have to see it, but now to really kickstart the business, think that’s a very good way to do that.

Adam Sroka (16:36)
hundreds and like there are parallels to like my journey with hypercube started on my own and yeah it was very much like data science for energy companies that do trading affairs in Scotland because that was like super niche and I knew the people and then you learn your processes you get a bit smarter you learn to abstract away some of the challenges it’s definitely I think one of the better ways to start because you kind of learn as you go and

Stephan (17:01)
Exactly.

Exactly.

Adam Sroka (17:02)
and bring

your domain expertise and kind of get the ball rolling on things. So we talked a little bit about how the residential markets may be stagnating a little bit and we’re moving into C&I. There’s like huge growth in some of these areas. You talked about how almost every company in this space has to rely on Chinese manufacturers just because of way the economy has been set up.

Do you think that’s a risk to the European market? And do you like, what do you think European companies can do to kind of clear out advantages and efficiencies that will keep us competitive in space? Or does that not matter?

Stephan (17:36)
Right.

Well, I’m definitely not against Chinese companies. mean, we’re working in the best space. We work together with a Chinese supplier. I’ve been to China just two or three weeks ago to visit them. And it’s always amazing to visit China and go out, have dinner with our Chinese friends and so on. That’s the one thing I’ve worked, as I mentioned before, worked for Huawei, which

depending on your perspective as a famous or infamous company, but I think we all agree ⁓ it’s an amazing company when it comes to its technology expertise and know-how and so on. think I learned a lot. Also the kind of pragmatism of Chinese people towards any kind of challenge they are facing. I think this is something where we as Europeans could learn a lot. And that’s actually something I would like to bring into my Huawei experience. I would like to bring in my new company and also in this startup.

environment I’m now in. okay, anyway, coming back to your question, is it a danger? Well, it’s a challenge for sure. And we see it everywhere, right? I mean, originally I’m German. mean, the German car industry is facing huge challenges because of the Chinese EV manufacturer. So it’s definitely a challenge, but I think it’s not the right way to somehow block them out. So what can we do? I think it’s customer access.

So let’s face it, on the technology side, Chinese products are as good, maybe even better, depending on where we are. Maybe talk about battery cells. They are better than German or British or I don’t know, battery cells, if they exist actually. So On the technology side, Chinese products are as good as ours, if not better. So what we need to do, where we have to compete with them and where we can be better than them is custom access and adapting the products to our needs here. So this means

We speak the same language as our customers. And I mean this literally because we speak German, we speak English, we speak French, whatever the language is, right? And this in particular, in C&I business, that makes a difference. Maybe not in the utility scale business, that’s a bit of a different story, but then the C&I business, this is important, We we understand European requirements and regulations better. And I don’t mean from a purely legal perspective, but also the perception, the mentality.

Adam Sroka (19:34)
with you.

Stephan (19:51)
So when we talk about data privacy, cybersecurity, why is that so important for us? It’s not just because there’s a law saying we have to do this and we’re not allowed to do that. We understand the mentality and we can adapt to it. And then maybe this sounds a bit superficial, but I think it’s extremely important. It’s a European aesthetics, like the user interface. Like how do we interact with the device? Now in China, everybody uses WeChat.

Adam Sroka (20:13)
Yeah. ⁓

Stephan (20:18)
It’s a super powerful app. It’s amazing what you can do with it. But have you used WeChat before in your life? It’s a terrible user interface. It’s terrible. And I think for our Chinese friends and colleagues, they would think the same about WhatsApp or Signal, whatever you use as a messenger service, right? But it’s a different mentality and we can adapt or we have to adapt our systems to

Adam Sroka (20:33)
Yeah.

Stephan (20:43)
what we think is good as a user interface. maybe summing it up So customer first, listen to them, talk to them and act accordingly. And then the other thing is when you talk about the intelligence of the system, like the energy management system, at least the critical components like the ones which control the communication, but also the user interface that should come from Mars from a European supplier.

Adam Sroka (20:56)
you

Yeah, that’s interesting. you touched on a really good point that I think techies, especially a really bad at where you like there’s a book, a textbook or some sort of course or Y Combinator or Stanford Uni have said this is the way you do modern application design. And this gospel and everyone should believe in it, blah, blah. And that just totally disregards like the human element and cultural significance of just like

Stephan (21:32)
True.

True.

Adam Sroka (21:33)
other part

of the world. It’s just like this one school of thought. Yes, a lot of us have indoctrinated into that school of thought. We do like the shiny Apple interface and Google Chrome and all that, but actually there are other ways of doing things. And some people might. And yeah, you lose that local specificity, I guess, if you’re trying to be too global and too generic or too Western and stuff. I wanted to touch on the idea around

Stephan (21:46)
Exactly. Yes.

Adam Sroka (22:00)
potential spin outs and structural changes within our city holding. Now, without breaking any confidences, what’s the strategic thinking behind building really focused businesses all sourcing under one umbrella?

Stephan (22:12)
Yeah, so I think I alluded to it a little bit before. We started the C&I business on the same company, right? That was RCT Power. And I think from a strategic point of view, it’s real customer access and learning from the customer. So in the old RCT Power, we have an established product. We have an established approach to market via installers.

we have customers basically. So it’s established business which works pretty well, sometimes better, sometimes less good because of, as we said, the market is not growing, but in principle, we know how to deal with that market. Now in the C&I business, we have new ideas, we are moving fast, we’re breaking things maybe to some extent. We’re discussing a lot, we have pilot projects,

we have some kind of lab where we play around with the system and see what works and we break things. So we were moving fast, it’s this kind of startup atmosphere and obviously we put some money into it and now we’re getting money out back. And so it’s completely different to the established business we had in the residential sector. And I think…

both sides of people working at the residential sector, but also the people working in the C &I business. They felt like we were working in completely different areas, segments, business areas. I don’t know how you want to call that. And this caused tension. Let’s face it. This caused, we are still friends, but it caused tension. so at some point we thought, okay, we can probably work more efficiently if we just separate these things and have a clear cut.

We can still collaborate on certain aspects, but then we do it as two independent companies and let’s try that out. And that’s what we did. We’re still in the same building. So we are still having lunch together if you want, and we’re still talking, but from a administrative point of view, we’re like to separate companies now.

Adam Sroka (23:59)
I think it’s a really smart approach and like cheesy, but it’s actually the reason our business is called Hypercube. I wanted to build a really, there’s a few reasons. It was also one of the first algorithms I’ve learned was Latin Hypercube sampling. I wanted to build a really focused like energy sector specific consultancy and really build deep domain expertise. And I always thought if I ever wanted to approach another vertical, I would spin it up as a subsidiary with a different brand.

Stephan (24:06)
Uh-huh.

Adam Sroka (24:27)
the end dimensional nature of things like that way. So that’s where the kind of aim converged on a few things. But I think it’s a right approach because you want to, again, like I said earlier in the conversation, in our experience, residential versus C&I versus utility scale feel like totally different worlds, even though you’re still managing batteries or virtual batteries and things like that. it’s a really interesting approach.

Stephan (24:30)
Yeah.

Exactly.

Exactly.

Yep. Yep.

Adam Sroka (24:53)
So you also lecture on energy and digitalization. That’s a very interesting topic that I’d love to hear more about. What do you see as the biggest risks in the digital layer for the energy transition across the

Stephan (24:57)
You

Yes, that’s true. I’m a lecturer at the University of Freiburg and also in Southwestern Germany, actually. I started this topic maybe, I don’t know, five or six years ago, and it’s a bit of yeah, the title says a lot and nothing at the same time. With all these AI and large language models and the associated energy consumption, it became quite a hot topic. So I’m quite happy to do that. And actually just,

yesterday I committed myself again for next year. So any of the listeners who might be interested in this topic, you’re cordially invited to join my lecture, but it’s in person. It’s not online. I should mention that. But coming back to your question, I think there are two sides to the coin and we have to make sure that everybody knows which side we’re talking about, right? So one side is how digital technologies can help us in the energy sector. And I think that’s

really where your expertise is. Just to give one example, The modern energy grid or electricity grid will become much more complex to control because of all the many different energy vectors we have. We have electricity, have natural gas, we still have in Germany coal power, nuclear power, know we have battery energy storage, and so on So we have different energy vectors, we have the volatility of the renewable energies,

We have a lot of power electronics, but at the same time, we have less and less inertia, rotational inertia of gas turbines and steam turbines in the grid, which makes things more complicated control. And the digital technologies is not an advantage to have them, it’s actually the enabler for the energy transition. So without modern technologies, without modern control approaches to the grid, to the energy grid, to the electricity grid, we will not be able to have.

to manage the energy. So it’s an enabler. It’s a must have. So that’s one side. The other side of the coin is that there’s an increase in energy consumption due to digital technologies. And buzzwords here are, for example, blockchain technology, particularly the proof of work scheme, which we use in Bitcoin, but also obviously the training and inference of modern AI models, like in particular, the large language models, like the ones for OpenAI, Anthropic, or Mistral here in Europe.

There’s certainly a risk that the demand due to generative AI, to agentic AI and all these kinds of new applications may dominate the electricity consumption pretty soon. So that’s the danger. And to come back to my lecture, deals with that latter part. It’s not using digital technology to enable the energy transition, but really understanding from the fundamentals, like how does the transistor work?

How does the computing architecture, which is so-called von Neumann architecture, why is this a disadvantage from an energy perspective? We’re shuffling back and forth a lot of data and these kinds of things. That’s what I talk about in my lecture.

Adam Sroka (28:04)
That’s really interesting. It’s a space where I think the social numbers recently that said, like, if you look at like the compound annual growth rate of the energy storage infrastructure specifically, it’s like sort of 15 to 20 % across Europe, which is really good, right? It’s quite a good number. Nice, nice chunky market with billions of investment, like tens of billions. But then if you look at the software for like energy,

Stephan (28:23)
Yes.

Adam Sroka (28:32)
AI and energy, the CAGR is closer to like 30 to 40%. It’s just so much that shows that everyone’s cottoning onto the fact that, yeah, you’ve got these assets, but how do we manage them? We can’t just turn the dial like we did in the gas power station and get everything we need. It’s a really interesting space. We had, you touched on data, like AI and the power demand. We had Giles Hanglin from Apertura on a while back and they’re doing really interesting stuff around.

Stephan (28:38)
really?

Right, right, right. ⁓

Adam Sroka (29:03)
co-location of beds and new data centers and some of the clever things you can do. But I’m interested, like you’ve talked about that a little bit as well. How do you think sort of the digital infrastructure and the energy storage networks can sort of work together more intelligently?

Stephan (29:17)
Yeah, think that’s that’s I haven’t heard of this of what he’s trying to do. So I would be I need to go back to your podcast and listen to him because I think this is potentially very interesting business case also for us, obviously. I mean, there’s the straightforward use case that you have a data center, you have a diesel generator as a backup, which kicks in if there’s a blackout and these kind of things. But it takes I don’t know, something like 60 to 90 seconds to wrap it up. So you have

you deal with these 60 to 90 seconds and that’s where the battery energy storage comes in already today and has been there already for many years or decades even. Back then it was lead acid batteries, now we use lithium ion batteries. So once you have the battery there and it’s typically depending on the data center it can be multi megawatt systems, right, so pretty big system. Once it’s there and you have paid for it anyway you can use it to participate in the electricity market

if that’s possible, other things, right? So that’s the straightforward thing to do. But what I’ve thought about a little bit with former colleagues of mine is really, and I think this is it gets really interesting to make the data center actually from a grid perspective smarter. In other words, to make the data center flexible.

Now we have in data centers, it depends on the kind of data centers, it will not work for any kind of data center. But in the data center you have urgent compute tasks, for example, you have a Google search, you don’t want to wait for your answer to your Google search for 10 minutes or an hour. So that’s a time sensitive task. But training runs for large language models, for example, they are not time sensitive. And…

they are the tasks that are in between. you need to have, what I would like to see is to have an energy management system for data centers, which can really prioritize compute tasks, shift them temporarily, but also geographically potentially. Google runs data centers or operates data centers all over the world. So maybe you can do a certain task somewhere in United States or in Ireland or I don’t know where, right?

So can shift them also geographically depending on whether you have a lot of wind power here or a lot of solar power there and so on. And Google is actually an interesting example because they are already doing that to some extent. However, you need to buffer still between the grid and the data center. And that’s where you need the battery energy storage system to smooth out some fluctuations from the grid, but also when you have unexpected compute tasks, which are very urgent.

because the energy management system for the data center will operate on, I don’t know, 10, 15 minutes time scales, but computers will be coming in much faster, for example. So you need a battery energy storage system to smooth out these kind of fluctuations and give you additional flexibility in operation. And there are some startup companies doing something like that.

Let’s see if RCT Pro is successful and we have the resources to look into this in maybe two to three years. I would be very excited to do that.

Adam Sroka (32:26)
Yeah, I wish you all the best in convincing Google to take you on board for their new build. ⁓

Stephan (32:28)
Yeah, it’s a fancy

topic. It brings together everything I’ve done in my previous career from ABB, but also Huawei and now a startup company where you can think about these fancy topics and so on. So let’s see, but it’s an exciting area.

Adam Sroka (32:37)
minute.

Being in Scotland, we can’t not mention wind. We love talking about how windy it is. And you’ll be aware of most grids, right? There’s huge challenges around high voltage, like the skills in the high voltage space and getting access to transformers and stuff. And so actually with Scotland, we produce loads and loads of wind and yeah, we pay for curtailment and everyone’s furious about that.

where there’s all the green power we’re paying to shut off. we’re talking to someone about, as a global economy, we’re now on the rampage to build loads of new data centers. And actually we’ll point around non-urgent compute tasks. like you get into chat GPT or Gemini or whatever, and you run your deep research task, well, it takes 15 minutes. So that could be done at a data center in the highlands of Scotland, right next to a load of windfall, they’re always getting turned off.

Stephan (33:35)
Yes.

Adam Sroka (33:37)
and actually drive down the cost of the compute there. It’s not going to fix it, it’s still going to create more load on the system, but it’s going to smooth out where the new demand comes from. I think the OpenAI data centers, no one cares about sub 500 millisecond response from chat GPT, right? They’re never that quick. So why don’t we dump all them in the places where it’s really, really sunny, it’s really, really windy. We’ve got good natural resources and access.

Stephan (33:48)
Exactly.

Yep.

Adam Sroka (34:07)
That doesn’t need to be in the centre of London or just on the outskirts of London. It’s how to get really really good cycle times and stuff. I’m interested to see how it plays out and what countries, so again Giles was talking about how he’s trying to encourage the Scottish government to really lean into the kind of data centre boom in the same way Dublin and Ireland did like 20 years ago. But time will tell, we’ll see.

Stephan (34:10)
Exactly.

Time will tell,

yeah. I think you put it much nicer than I did, but that’s basically the idea behind it, Latency, you don’t need latency in these tasks. For other things you need it, but latency is not an issue here. And then you can, that brings you a lot of flexibility. ⁓

Adam Sroka (34:35)
Yeah.

It’s funny, like you watched like technological evolution and for like 20 years we’ve just been obsessed since the internet was started. We’ve been obsessed with making it as fast as possible. And then new technology comes through and all of a sudden no one cares anymore. Wow, it’s fun. Yeah, we don’t care. I it’s really clever and it looks like magic. I wanted to talk a little bit about that you’ve mentioned being, how do you build a digital first energy hardware company? Like I think that is such an interesting topic. What does it mean?

Stephan (34:56)
Right.

you

Adam Sroka (35:14)
Like how close to how possible is it? What are your thoughts?

Stephan (35:16)
Yeah,

that’s a bit of a dangerous question because you’re the expert here. So I’ll try to answer this question from my perspective, but I’ll hand the question back to you afterwards. So yeah, I try to keep my answer short and then I’ll ask you. I think there are two aspects to it, maybe even more, but at least two and that’s how we do it in RCT Pro. So first of all, there’s a digital…

end-to-end customer journey. So this means from the very first contact we have between a customer and us, we track everything digitally until the very end. So until we install the system at the customer’s side and then also provide services, maintenance agreements and so on. So we track that, we have a digital customer journey. It’s not only that we track it for ourselves, also the customer.

as a account on our website where he or she can track it, where we are in which phase and so on. we don’t have it yet, but we’re working on that also offering rather simple tools for the customer so that he or she can get the feeling like, how do I size my best system? What is a TCO? What are typical payback times and these kinds of things.

and later on obviously having battery analytics and so on so that the customer can also see these kind of things. So that’s the digital customer and to the end journey. That’s the one thing. Then on the more technical side, we are building something which we call the currently the RCT operating system. It’s not a super fancy name maybe, but it somehow describes it pretty well what we would like to do. So initially, this will be a kind of

Let’s call it a site, EMS, site energy management system. But it’s really intended to evolve into like a digital layer. You mentioned this word digital layer before, but yeah, it’s intended to evolve into a digital layer which interfaces to third parties like a private cloud. So we would like to work to have a private cloud, not a public cloud for well mentality reasons, these kinds of things. But anyway,

which interface to the third parties like the private cloud to service providers for battery analytics, for energy trading and so on. But again, also to the more fundamental software components like the energy management system running on the battery energy system, which then control the battery management system and the inverters and so on. So I think the latter part is pretty close to what you or what HyperCube does.

Is that correct? Yeah, what’s your perception? What is a digital first hardware company? So I’m quite curious to hear your answer to that.

Adam Sroka (37:43)
Yeah.

So what HyperCube does is something that we should, yeah, we struggle to communicate sometimes, but I think actually in the context of this conversation, we’re almost the staircase to get you where you are to digital first. think like that’s the point of HyperCube is that we can come in, we can deploy our tools, our systems, our expertise to help take you from wherever you are. Maybe you’re nearly at the top step or maybe you’re right at the bottom and start building up towards being

digital centric or digital first. I think digital first is like, maybe it’s just semantics, but digital first is really difficult in a space where there are no digital centric businesses. Like there aren’t hardware businesses that are totally digital. So there’s almost like, to get on that journey, you have to innovate the way this is done. Cause no one else is doing it right. So

So the mentality is right, but actually the kind of feasibility almost isn’t there. can’t just adopt digital processes because you’re going to have to invent them. And I think the only way to do that is to look at, how do you run a normal hardware company and then look at every process, every bottleneck and think why is a human doing this job? Why does a human do this? And like the only acceptable reason in a digital first company is that

There’s specifics around human expertise, the context or the complexity that we cannot automate away. But copying data from A to B or like checking files or running numbers or building reports, that is all things that computers can do faster, better, more accurately, like with less risk and better security and all that, better scalability. I think the other lens to look at is…

Are there fundamental roles that are intractable or couldn’t be done by humans? ⁓ I find this hard to explain without an example. So the one I’ll use is we did a body of work with a really cool star called Gable, and we’re to market with their product in the UK for energy companies. And it’s really clever. It scans code bases and it scans data assets within a cloud environment.

and then it can draw up the graph of every data asset and where it was produced and how. And this is really important for maybe businesses that have been going 20 or 30 years and maybe have a CRM with like 15 year old data in it that was produced by a piece of code that was deleted seven years ago. Right, like how do you find where that number came from? It’s really hard. And actually that was, that probably was possible.

You could probably could do that for a million transactions. You probably could do that two years ago or five years ago, but it would have taken you thousands of humans, thousands of hours. Now the whole thing runs in 90 seconds. It’s a job that AI can do at scale that is just intractable. And I think there’s a lack of imagination across the whole tech sector about what AI can do. And a lot of it is really good, intelligent automation, business process stuff. Great.

Stephan (40:44)
Hmm. Hmm.

All right.

Adam Sroka (40:55)
But what about just fundamentally changing the roles that we could do? Like read every line of code we’ve ever written and read every transaction we’ve ever done in the database. Marry them up. Like it’s not complicated, but you couldn’t do it without these tools. So I think digital first, have to kind of have that second lens of like, is there a way to blow open this whole business process to deliver the outcomes

that we need. I always think when I’m talking about AI automation data strategy, I think like you want your humans focusing on the value add their unique expertise and the needs of other humans. And you want robots to do the rest, right?

Stephan (41:34)
Right, so I think from my career perspective, so I started, as I said, after my PhD, I started working for ABB and later on with Huawei. Both companies are intrinsically hardware companies. Like Huawei is not super old, but ABB goes back to the mid-19th century, right? For these companies, and I appreciate that because I like to

touch things as well. So I’m a hardware guy to some extent as well. So I appreciate that. But for those companies, digitalization is something you do at the end. You think about a product, what is the functionality? Like you have a transformer, it should transform voltages. You have a switch gear, it switch current or voltages, right? You define the functionality, then you do the engineering process. And then at the end, you have two more tasks to do. One is,

thermal management. And the other thing is, can we put some digital layer on top? Or something like that. And Huawei, okay, I’m a little bit exaggerating here, but Huawei, again, is also a hardware company, extremely good at hardware, but then digital, it’s not this digital first thinking. Now, when we build a new company, this RCT Pro, what we would like to do, and that’s, it’s a mindset, and you alluded to that, right? Any kind of

process, project management, customer journey, everything. We have to think about can we automate it? there digital processes or are there digital tools which make the process more efficient? Then how can we integrate it into our company somehow? So I think for me, when I say it’s a digital first company, I mentioned the customer journey and also the RCT operating system. And I believe in these kinds of things.

but the most important thing is really the mindset. And that is in some sense independent of a software company or a hardware company.

Adam Sroka (43:24)
Yeah, no, I totally agree. it is a journey and like, we’ll get better over and then the next generation of businesses will be better and so on and so forth. And like, so yeah, like funny sort of time, like I am, I was laughing, people talk about like, is AI gonna come and take my job? Because I think, well, yes, but can you first write down everything in your job? Write down, write down every task, how to do it in your job and then yes, AI.

Stephan (43:32)
Yeah, hopefully, yes.

Adam Sroka (43:49)
take it and like that’s a ridiculous task to do and myself like building stuff and I’m building like pet projects on the go at the moment and I I banjaxed all of my credits on like my coding agent and I was like have I done that I was really surprised I destroyed the credits that I had and I was looking at things and was trying to was like okay I’m not going to limit I went through and I found out that in the process that I was building

I was taking Excel files, converting them to Markdown and doing some analysis on them as part of a really complicated data pipeline. And I spotted that the Excel file that I was receiving, someone had formatted to the very bottom of the sheet, this would Y. And when it got converted to Markdown by the Markit Down agent, it turned it into a 32,000 line.

table of nans, there was just so, but they were written into text and then all of those lines were getting passed as tokens into the agents. And I’m like, there it is. That’s how I’m destroyed by limit. Because it’s just not the thing you think like a human, would just cut the table out or whatever, because I was kind of sort of all making the process. So I just think that, yeah, there’s so many hurdles and mistakes and you have to, say to businesses all the time, get used to making the mistakes. Like you’ve got to understand how to run through these hurdles and build up the muscle memory.

Stephan (44:58)
Right, right.

Exactly.

Adam Sroka (45:11)
of how

do you fix this and how do you get past that. Yes, chat GPT is probably going to like add your pilot project as a feature in six months. That will probably happen because they’re going after everything. But that doesn’t mean you should just wait until everyone else is in front of you. like you should still try and learn and mature your adoption of these processes. I mean, I’m talking of process and like, yeah, you’re a CTO of a really successful hardware company. You’re starting a new business. You’ve done lots of IT.

what does like operational excellence look like in these contexts?

Stephan (45:43)
Yeah, so that’s a good question, obviously. So I’m as a CTO, for me, the most important part is R &D. So when we talk about operational excellence in R &D, as a startup in particular, because it’s basically development work, we don’t do research, right? It’s really fast and focused innovation. So validate what our customers need, actually, through partnerships, through field installation, pilot projects, and so on. So test that.

And if it doesn’t work, stop it quickly. So that’s when it comes to R &D. Manufacturing, I mean, we want to go towards manufacturing. we have to see how it works out, but we want to go towards system integration, these kinds of things. So, and maybe this also differentiated to our Chinese competitors to design for quality from the start.

Like Apple is really good at that, right? The first Apple product typically is a really good product. So designed for quality from the start, tight feedback loops between R &D production, but also from the experience we got to get from the field to catch any issues as early as possible and then improve continuously. And then when it comes to the customer journey, make it easy for our customers standardized or with also partners, interfaces, clear documentation,

dedicated technical support, these kinds of things. So that’s also part of this digital customer journey. And in this context about operation excellence and in a startup, I once heard somebody saying something like that, data-driven decisions and ruthless prioritization. So we’re a startup, we don’t have a lot of resources. So we have to focus the resources on what creates real value for the customer and which can be scaled.

So I think that’s what operation excellence for a start at least is about.

Adam Sroka (47:28)
I like it. like it. This has been a great conversation. Thank you very much. Before I let anyone leave, I always like to ask if people in the audience want to find out more about you or more about RCT Pro or RCT Power, or if there’s anything you want to promote or plug, where should we direct

Stephan (47:34)
Thank you.

Well, obviously you can go to the internet. So www.rct-pro.com. You can contact me directly on LinkedIn. Just drop me a note, contact me there. It will be a pleasure to reply to you.

If you’re close to Konstanz, Southwestern Germany, close to the Swiss border, let us know. I’m happy to invite you for a coffee, show you around. There’s not too much to see at this point, but this will change in the near future. And we also go to trade shows, conferences all over Europe. So I’m sure we will meet, we can meet there as well. So yeah, just contact me wherever you want.

Adam Sroka (48:25)
Well look, it’s been an absolute pleasure. Thank you very much and I look forward to seeing what RCT Pro will to in next 12, 24 months.

Stephan (48:33)
Thank you for having me, Adam. It was great fun. Thank you.

Adam Sroka (48:38)
Thank you once again for listening to the Hypercube Energy podcast. As always, if you have any feedback or comments or you want to get involved, please get in touch. And if you want to find out more about what’s happening in data and AI across the energy sector, please join the Beyond Energy community and I look forward to speaking to you soon.