Episode 17: Increasing Solar Panel Effectiveness

Hey listeners, welcome to SphereCast, a podcast all about technology, technology advice, technology inspiration, and how real entrepreneurs have used technology to build their businesses from the ground up. If you're wondering how technology can support your business goals, rest assured, our guests have been there and done that.

In 1883, the earliest solar panels began to take form. Small electrical cells that could convert light energy into electricity. Truly incredible. Yet, as the need for sustainable energy ever increases, solar adoption simply doesn't keep pace.

Why? Because they're expensive, and not just to install, but to maintain. That is until our guest this week came onto the scene, one who is using AI to take the pain out of solar upkeep.

Hello everyone, Luke here with SphereCast, and this week Xenia and I interview Ingmar Kruse. Ingmar is the CEO of SunSniffer, a technology that measures solar power module performance using artificial intelligence. Kruse's platform provides information on each module's power usage and whether the module needs to be replaced, all with the push of a button.

Prior to the advent of SunSniffer, this type of work was expensive, more expensive in fact than the panels themselves. However, through his work, Kruse hopes to take the pain out of solar energy and to ultimately encourage more widespread sustainable energy adoption.

Ingmar's journey into solar power was far from linear, and his story is incredibly interesting. In fact, Ingmar's earliest work was with color printing and scanning and its associated software. Ingmar's diverse experience fostered a certain out-of-the-box thinking that ultimately led him to his work today, work that is quite literally saving the world. On the podcast, Ingmar talks through his full story, gives details on how SunSniffer got off the ground, walks through his typical day, and so much more. So without further ado, here's our conversation with Ingmar Kruse.

Hi Ingmar, welcome to SphereCast. We're so thrilled to host you. Tell our listeners who or what inspired you to become an entrepreneur.

So, hi Xenia, hi Luke. Yes. Well, thanks for having me here. Well, what inspired me, it's quite a decent time ago, making a change, you know, being part of something big. And I've seen so many interesting technologies and somehow participating in the path of development and bringing out those products really is pushing me and pushed me to become an entrepreneur.

I want to ask you on that, like, you said this is what's motivating you. So, could you tell me a little bit about the you you introduced the first color scanner in the market. So, I have not, I mean, I researched, but I are we talking to the whole world, or to a certain market in the US, which is normally the first market which gets all the, you know, breaking tech, so can you tell me a little bit more there?

Oh, yeah. Well, actually that's right. You know, your world is really ending where you can still see, or though it gets much further. No, it's more for the European market. And the US market on some areas, they really were early adopters, but not necessarily in the early days. The professional use of color scanners, that's when they were introduced, was actually more here in Europe as far as I can see it for that time. And so it is, yes, Europe.

Okay, great. So, and and what made you bring that to the Was it when you were in Atlanta for education and you I mean, I'm just trying to understand, connect the dots here, so that's what

Yeah, oh, well, that's that's I need to really roll back. That, you know, that was all driven from a really exciting development. I mean, there were so many disruptive technologies, and the interesting thing is everybody saw that it is disruptive. That's different today.

You know, we had, in the older days, where you used scanners, you didn't have an an easy way to to get an image printed. That was a really complicated and very expensive, time-consuming way. And if you introduce something, put a scanner on your desktop, where everybody's saying you have your mobile phone, at that time you needed scanners which were costing 500, 700,000 euros. So, we're talking serious money.

And now we came with this little teeny little scanners for 30,000 euros, which is today, well, that's expensive, but was super cheap. And you know, and you just put the picture on top, and you scanned it, and then you still had only RGB. And then you had to to make the conversion between RGB to CMYK. That's the colors you really print. And you know, it was exciting. You could see people, they never saw 24-bit images, and they saw on screen, I mean, when you see it the first time such a new thing, people are amazed. They were, you know, people just couldn't go away anymore from the computer. So, this was so much excitement in there, you just want to be in there. And this excitement came from everybody who saw it in the first place.

And I have to admit this was kind of a little easier to motivate yourself because, you know, you fire it up, and everybody sees and says, "Wow, this is where it goes." Today, that's not that easy. You show something where you know this is going to be big, but the rest of the world is saying, "What is he talking about? I don't even understand the business model behind." And the complexity at that point got much more and bigger than it was in in the past. But anyway, that was that motivated me that time, and there were other technologies where it was so obvious. But on the other hand, you needed It was much you know, the hardware was much more costly. And to start it as a student was really difficult.

You know, there is maybe a funny story. I remember when I was a student at that time, and I saw an ad, and I saw something from from Adobe. And Adobe, in the first time, they they, you know, they had PostScript. That's the first thing they had, and they licensed it to a printer manufacturer as the LaserWriter from Apple. And I saw that, and I saw the outcome, and I was so fascinated, and I was here in Germany, I was calling those guys up. And I said, "Listen, I have to meet you guys. I I have to meet you." So, I went there to to Silicon Valley, and I was sitting together with the founder of Adobe. There were a couple founders. It was John Warnock and his secretary, and we were sitting at the table, and I was saying, "I love this technology. I need to be part of it. Doesn't matter what it takes." Well, he said, "You know, a student, you know, very ambitious, nice. But we're talking you need to be a printer manufacturer because, you know." There was no room for me really. So, I said, "Well, but I still love it." Funny is that John Warnock still remembered me always when I was coming to the show, to the trade shows, and he was saying, "Hey, this is Ingmar! He really believed in us." And they were already big, and they got bigger and bigger, and you know. And everybody always thought, "Wow, this is a super important customer." No, I just loved what they did, and I still do.

So, anyway, it's it's, I guess, it's passion what you do. And if you're passionate about it, it's it's hopefully driving you into the right direction. At least it got me in a nice direction.

Thank you, that's a fascinating story. To follow up on that, you launched the first color management system, which is ColorSync, and then the color proof software, the first digital color simulation software back in the '90s. Could you tell us if that's how you landed the rare license opportunity from Apple for the production of Apple computers?

Oh no, that has something totally has nothing to do with that. No, there wasn't, well, you know, we went into this technology, as I said, you know, you need to change color models from RGB to CMYK, and I don't want to get your your people bored, but you needed really some serious mathematical computation stuff. And it was at that time it was that was tricky because limited memory, I mean, serious limitation.

We're not talking gigs, we're talking megabytes, you know. So, big difference, big difference, and you were really computation time was very little, and, you know, transforming multidimensional color matrices into another model was really challenging. So, yeah. I mean, color, I mean, if if you if we're used to 256 colors, it just fascinated you to go 24-bit and to do something with it. I mean, not only to have nice colors on screen, it's fascinating, you know, if you can get it printed. And to do really something productive with it, you know, change the world how it's working, making information quicker. Yes, remember there was a time before internet, so nobody had internet connections. So, you still had to communicate something, you needed to have something printed.

And that was awfully expensive. And, you know, and this is one process how to make it quicker: you got a picture, you scanned it in, you could make the color separation of it and send it directly to the printing company where they printed out the films, high high-res films, on their Linotypes and whatever typesetters. And yeah, and you saw, and it was printed the next day or the same day. And that was not possible before. That was introduction of desktop publishing and, you know, 24-bit image processing. It was great. You saw something you had it before, and it got printed. Fascinating.

I agree. I think just moving from Okay, I think I'm giving my age away, but just just going from, you know, black and white printing to color was just an amazing jump. I I I remember, you know, getting the color printer, and I think that's quite a few years ahead of what you're talking about, so just getting a color printer and being able to print a photo. And I used to buy these photo, you know, paper just to print it off to try out because the concept was that you have to get to a studio to get these, you know, reels developed and stuff like that into photos. And when we could do it ourselves with digital photography, I think I was I was like absolutely into that when I was young, so.

Yeah, you're right Luke. This was the same effect. It just was printing going to printing press, while this is as you said it was printed on your own printer. Same revolution. Of course, hadn't didn't have the scale for the masses, except that masses now were buying color printer. Yeah, it's it's it's fun. And it had nothing to do with the with the the clone business, that's what it was called, from which we had a deal with an a manufacturing place in Taiwan called UMAX. And they had an Asian deal with Apple in place. And where then we got an OEM deal where they were producing Storm, that was Storm computers, with our name on it. And it was an an Mac OS, you know, System 7 at that time, if I'm not forgetting. And we were really competing then later on one-to-one against Apple. Steve Jobs was at that time not there. And we were really competing against Apple computer.

I mean, today it's like unbelievable. What did he do there? And they were but I'm telling you one thing: the message there was always the same. It was great. I loved it. And, you know, I still have an Mac Macintosh SE/30 here next to me sitting, looking at me. Actually, I I had it somewhere, and I put it back because I still remember it so well. That's the first 24-bit I have a 24-bit color board in there.

It's a collectible, it's a collectible now, definitely.

Yeah, it is, but I it's more collectible for me than for anybody else, so. It's sometimes fun.

You know, I know a person on YouTube I follow, he actually he's so fascinated with the iPhones and everything Apple, his channel is called EverythingApplePro. So, he he opened a cafe in the US, and he's got like everything made by Apple on the wall, literally, from old phones to old computers and everything. So, yeah, it is some kind of a business nowadays as well.

Oh, yeah, yeah. Well, you know, I started my first Apple was an Apple II Europlus, so some years before of that under CP/M. So, we had Anyway, I don't want to get the younger crowd really bored of this old stuff, you know. The new computers are I mean, every every every decade has its fun, and you remember it, and some remember then VT20s and so on. But this is now, now we're talking graphic boards and superspeed, and how you do CUDA on it, and, you know, reuse them for AI stuff. And the fun is is still going on, and the the great thing is everybody can do it, or at least, you know, with a certain pricing level you many people can do it. And the capabilities are are great, and it's getting better and better.

Ingmar, I'd like to ask you about your current venture. Could you share with us, what is SunSniffer, and why did you create this service?

Oh, yes. Right, right, right. You know, what we do, we digitize solar. And as you know, you know, there are more and more solar modules on the rooftops and in the field. And actually, solar by itself is very disruptive. It's in super disruptive industry, because pricing were driven down from Energy production through solar used to be 10 years ago something super expensive, and it's only would be affordable if the government is is is helping on the feed-in tariffs and, you know, helping to grow. Well, that has totally changed. And solar got so cheap that today the electricity produced by solar is the cheapest energy you can produce. And this is incredible because it's CO2-free. Well, almost free.

And though actually 2001, 2002, when I went out of the Apple business, because there was only software business there and it was so I thought it's I'm I'm good for a change. Actually, it was 2003, something like this, I went into solar, because I thought it's it will make an a change in this industry. At least it has a perspective. Nobody could foresee that dramatic price reduction, but it's just great. And anyway, long story short, it got real cheap today, but there's still a big issue: what needs to be resolved?

Today, there are a lot of monitoring systems and people saying, "Well, I can see what my solar plant is doing." Yeah, right, they can really see what it is how much it is producing, but it doesn't tell you if it is producing exactly what it should produce. And that is actually for an investor really important. Are you sure that what you're producing there is is efficient? Can you maybe get 10% more outcome? And that's where people start to struggle with.

And there is one interesting calculation: if you see from your monitoring system, and this is interesting now, if you see there's an issue, you still need to send today somebody to the plant. And the point is a module today, a solar module, solar panel, costs, let's say, 60 bucks. And just to investigate that panel, it's not replacing it, costs like 100 bucks. Okay? So, it is more expensive to look at the panel than to replace it. So, what's happening today is people just live with degraded panels, because it's just too expensive to touch them. And that's where we are coming in.

And we developed a technology, the sensors super low-cost, because, you know, solar needs to be cheap because of the disruption, remember? So, we developed a sensor which you can which is super low-cost. You put just in each module, has a little operating system, has all, you know. Actually, it's as powerful as a VC-20 at that time. So anyway, it's in each each panel, and we measure the power of this module and communicate it via power-line communication. So, you didn't need any additional cable. So, we send that information, and then we collect it again, and then we do artificial intelligence and can tell you how good the module is, if it's soiled, what's the quality level, and if it is within the warranty or if it's degrading. So, modules every year degrade in power, okay? So, this year is 100%, next year is 98%, and so on. And some modules because of certain things degrading quicker. And you know, is it within warranty? Is it out of warranty? So, we're checking this all.

So, what we do, we're not taking only the measurements in, that's everybody does, but we take irradiation level, all in it, calculate it through an AI model and and a digital twin, and see how much the the plant or the module will do, and we classify modules in loss classes. So, long story short: we have in technology, super low-cost sensor, and we can tell you how good each module on the roof, or in we're now addressing this for very large fields, how good they are.

So, then you press a button, and then you know exactly those modules are out of warranty, they are still good, they are running into risk, or, which is even a bigger problem is, how much you lose due to soiling. Because due to soiling, you can easily lose 20% of power, easy. And and that's an information you need to know. And the modules are not soiling equally. So, some modules are soiling 10% more. Now, how do you deal with this? And it shows you you need computation. You need artificial intelligence.

And we're actually we do not have really very much competition in this, and we can tell you exactly how good those modules are. Besides this, we have developed another hardware where we can shut down modules and optimize modules as well. So, we build a hardware which you can use even in smaller plants. And that's that's what we are offering to to OEMs or to big module manufacturers. And the technology, the AI technology is for for big companies who are having huge amount of solar plants in their assets.

In terms of hardware, I think things evolve every year, so like you said, soiling is is one of the biggest issues with the the panels. So, for the audience, just to clarify, soiling soiling, I guess, is the accumulation of dust that's building up on your panels over time, and it's it's making the light not, you know, go in properly into the panel, and it's just, you know, deteriorating the performance. So, so obviously, the panels have been evolving. The software has been evolving. So, what what I think we would love to understand is, in terms of when you, you know, said, "Okay, we'll do a proof of concept or an MVP," what was your thinking behind it? Did you go for something that was proven, and you know, spent time on, or did you go for perhaps something that was cheap and dirty just to, you know, try something out, let's prove to the market we can do it, and then we will go in with the next, you know, version?

Well, that's a good question. Actually, at the beginning, I really didn't know what we are doing. Because I looked at that, and I said, "Hmm." I mean, we have different ways to look at. Is there a differentiation for a marketing-driven opportunity, or is there a technical solution to it? And if you go a technical solution that you will define something at the end, you need to be really totally in it and to understand it. But who knows, and who understands it at the beginning? Not us, at least.

So, we just did it. We said, "Well, let's put up a budget. Let's develop something small, and let's watch it. Let's see what we're getting out. What information? How can we use it?" We just did it. We said I saw there is no digitalization right now, and I just know this is now some years ago, over 10 years, I just said, "Well, there's a clear path, no way where it's going, but the path will be in good, you know, will be in a good direction. Might take a little longer." So, that's that's the path we did. So, we developed a sensor, we started software on top, and, and, and, did all kind of mistakes you can do. And yeah, and and we learned a lot because well, the issue is that you have a lot of readings, a lot of readings. And, you know, to use all the great calculations out there, you need a a certain way how to manage your data.

And that is not that old that you can, you know, manage huge amounts of data for little money. I mean, there were some databases, but they were quite expensive, and we really like to go the path for open source technology. And so, from first, we had a monolithic system, of course. Now, this is getting more technical. I'm not sure if if if you guys are

No, we go we go fully tech, don't worry about it.

Okay. So, we had really a nice, you know, MySQL, you know, database. And, you know, the issue is to to scale MySQL is challenging for smaller teams. And then the analytics, you know, they're drawing all huge data. And well, we had to go through this big process, you know, having the honeypot there, the nice part of analytics, which is for everybody greatly available, requires a certain infrastructure, or it's not going to happen to you. So, this was kind of frustrating. So, I think that a lot of companies, at least out of my perspective, cannot even use this great technology because they don't have the proper infrastructure on it.

So, we had to, you know, go from monolithic into microservice-driven, event-driven infrastructure. You know, the whole world talks about it, I'm not sure how many are really doing it. So, and that was really a hard way for us because the skill set of the people were totally different. I mean, you had people who had, you know, come from MySQL, Postgres, they are not necessarily knowledgeable for big data architecture. So, that was really challenging, and took us a while, I have to admit.

So, it's it's Microsoft mainly, I'm I'm guessing, because with MySQL, you you would have worked with Azure.

Oh, no no no no. We host everything on our own server. Even Azure and Amazon is is just getting too expensive. Right. It's it's not working it doesn't work out. It's it's No.

Okay. So, so on-prem then, is it?

Hm?

Is it on-prem? Okay. Wow.

Yeah. Yeah, yeah, yeah. So we uh well, remember, solar is really low-cost. You need to compete in this. And as they are saying, "Hey, you make a lot of money putting it in our cloud," but well, first, people need to see they're not used to pay a lot there. So, it's it's an it's a process. It will go in a good direction if they see they get some more out, and then there's room for maybe higher hosting costs. But so far, we have to do that by ourselves. Of course, it's easier for somebody just using it.

Yeah, yeah. I mean, I found in some instances I was speaking with one of our That's for another podcast, so I spoke with the CTO, and they are utilizing Google Maps API. So, Google are actually charging more once they got above a certain limit, so your rate actually go up. So, which sounds very strange considering it should be less since you're using more volume, but they've increased the rate. So yes, that's the trend. So.

Well, you know, it's it's pretty fair because I guess they're not charging more, they're just charging what it should be charged for. And they give people the the opportunity to get in, and if they really generate business, then they can even afford it. And so I think that's probably the logic behind it. But I guess those guys have a better idea on on on the businesses models than than us. And, you know, usually you have nice in the beginning, everything is reasonable, and if you need more service, when you grow, it gets really expensive.

I see. So that CTO I'm talking about, he decided they decided to go in the direction where they will develop their own mapping system, so they develop their maps from open-source systems, and now they use their own, you know, data from the fleet that that's running in on the streets of London, so, anyhow. Okay, that's that's for a different day. Okay, thank you for that. That was that was insightful. Xenia, what's next?

Next, I have a COVID question. Ingmar, this is a the staple in our podcast now. So, COVID-19 has impacted most of the businesses, as we know, across the globe. How has it changed the way that SunSniffer operates? Did you have to pivot or explore new markets?

Well, yes and no. On the one hand, it actually accelerated our business because COVID had the reaction that, you know, the borders were closed and you couldn't send people in between the countries. So, to have in remote a view, and to have a deep look into your plant, is really is really helping people of, you know, that you don't have to send somebody on-site. I mean, if cost if you don't pay too much attention to cost, unless you put it all the costs at the end of the year together, you know, people, I mean, if you can drive there. But now they didn't have a chance, so they need to dig into something new, and that was good.

So, from the term from the from from the development level, I mean, developers we are spread it all over the world anyway, because solar is all over the world. So, that doesn't really make a big a big difference to us. But what was good actually, and personally I hate it because I see that there is a need for personal contacts, beside just communicating what the tech specs are, you want to have some personal contacts. And that's, of course, as we probably all feel it, is is really missing, and it's really bad.

But I could participate, and my team members, on many more conferences, which wouldn't be digitized, and, you know, where I just quickly see the sessions which are interesting for me. And due to COVID, they made it reasonable to to participate and to watch them. So, actually I was looking I was seeing many more conferences, and only the parts I really was interested in. So, that was the positive part about this this really bad disease, well, it's more than a flu. But so, you know, it has all its opportunities behind it, and that was quite positive, besides, you know, the personal contacts we all didn't have. And you know, it's not me, I guess, we all are there.

So, we didn't, of course, the opportunity out of the lost contact information, that's kind of difficult to to calculate. But anyway. So, actually we did quite okay in COVID, and ah, wish only it's over soon.

Yeah. Absolutely.

Yeah.

Okay, a bit personal now. So, Ingmar, would love to know, I could say like you're you're an innovator, right? And that takes a lot of balancing act, I guess, in life. So, you have family, you have work, etc. etc., all all the things happening. How do you keep on top of, you know, what you do in business? Could you describe your typical day, you know, what it looks like? Your routine like, I'd love to understand. Like, do you get up at 3:00 in the morning? Maybe some of the people we spoke with in the past get up at 3:00.

No. Yeah, well, actually, yeah, it happens, especially when certain things are not going in the right direction. You know, if you're passionate, which is nice on the one side, but on the other hand, of course, you take it serious. And obviously, that didn't happen when I was 25. That's with the later years. I've seen this more, that you're really at a certain point you can wake up at 3:00 in the morning and then, you know, like, "Ooh, how do I manage this? How can we get it going?" and you know. But that shouldn't be the the normal way, but happens.

No, usually, my my day starts morning at 6:00, 6:30, and I do some work out, which, especially now here in Germany, all the fitness studios are closed, so, you know, usually I go three times a week to work out, and that's, of course, not possible. So, I do some home workouts, and I I got a bike because in the winter was really cold here, and if it's raining and 3 degrees, ah, so I got the bike. And in-house, and I work out on that at least, and do certain, you know, I have some some weights, I do a couple of things. Could do more, you know, could do more, I have to admit, but I do it every day.

Fair play, yeah. As long as there's more hits than misses, I guess, you know, it works.

Consistency, yep.

Yeah, could could be more, but, you know, usually I'm I'm running more and work out more, I mean, and and, you know, now it gets summer, you can get more outside, so this is all good. But it's not as easy as it sounds, "Yeah, wake up," and you know, "I have this, and you." I'm always struggling with myself, even when I do it on a daily base, and I say, "Oh, I have to do it. Yeah, ooh, isn't don't you have the coffee first, and that more chilling and take the newspaper, and sit outdoors?" Well, if you do that, you're, you know, you're not going to do it. So, you know, no, go back, do it.

Yeah. Well, I I I'm glad you said this, because I thought I was the only lazy person. But I'm not saying you're lazy, but like, I I don't love to exercise. It's like something I have to, because I know that it's going to help my you know, I think mental side is much more better, like you you feel better as a person throughout the day mentally if you exercise in the morning. So, do you do any meditation, any mindfulness? Are you into that?

No, I'm not. I know that a lot of people are really

Especially in Berlin, it's very popular.

Oh, I know, I know, I know. I'm sorry. No, I'm I'm not I'm No, I I did it sometimes in the past. I took some courses and there was was was I liked it. No doubt, it can get really powerful, and but I'm not really into that.

I know it is you should say it because the whole world does it, and now you're really trendy doing it. I believe in it. Don't get me wrong. I like it, but I'm not doing it.

I I I think, you know, anything could be meditative, so like, I work in the garden, and I I find it really meditative, because I'm fully focused on that. So, I it's it's whatever, you know, you can get you focused on. So maybe your passion for technology and, you know, solving problems is meditative for you, so you just have to do that. So, yeah. I mean, it's not everyone's cup of tea, I agree.

No no, don't get me wrong. I had a meditation course, my first one I think I was 11 or 12, I took. And I got a it was a very good trainer, and I still remember the sentence he told me I which I should tell myself, and I still do. So, I believe in all of this. I I do it on a certain way, but I'm I'm not celebrating it like people, you know, talk

Mainstream, it's very mainstream.

You know, I'm I'm just using it, and I know if I need meditation for myself at a certain level, but I don't celebrate saying, "Yeah, I'm doing it every," you know, "Om, om, om."

You're not bragging to the world that, "Hey, I'm meditating and I'm here."

It's it's it's not me. No.

So, we've come to the final part of our podcast. This is where the questions can get a bit tricky. What continues to be the biggest challenge from a tech perspective for you, Ingmar? And also, how do you wish technology would evolve to make your business more efficient, and what advice would you give to someone who's just starting out?

Yeah, changing our architecture into the microservice-driven, you know, event-driven services, it was really challenging because you really needed it new new guys on board. That was we didn't know whom we are getting. Do they really know their job? Challenging.

And just to follow up on that, how do you wish technology would evolve to make your business more efficient? And perhaps you can share an advice with someone who is just starting out.

I wouldn't I'm not giving you advices because it's I wouldn't do that. But you know, actually, I think we're in a good point. There's a lot of open-source stuff out there. The only thing is you need to get somehow a little more time to learn all of that. But I think it is it is a great time, because you just you see on YouTube great videos and explaining that, and it's even free of charge. So, there is is not the burden of a big entry for paying a lot of money for it. So, I think it's it's great. It's really that's it's happening.

So, can it be better? Yeah, well, I think it is I don't know. I just learned it always more difficult in the past, so I think it is great, and I love it. So, you know, when we're really at 130%.

I like that. I think, you know, I I really like the advice part from you, the answer to that. So, and I I'll give you a reason, so I I picked up, you know, a meditation course about 2 years ago, so I have like a guru. And the thing and and, I mean, I shouldn't say this on podcast, but I have not read any books since I I started doing meditation. The the books, I you know, I think literature, when you learn some technology, etc., is good for you, so I read those books, but I have stopped reading all the pop culture, psychology books, how to do this, 11 things, 12 rules, blah blah blah.

And and that's because you cannot learn if you really learn in life if you are constrained in these boundaries of books that you're reading. So, you have to be really at a point where you say, "I don't know anything," and you discover everything yourself, take a challenge and try and solve it instead of putting walls around you that only these things are true and these things you can do and cannot do. So, so I think in the last 2 years, I've I've done that, and anyone who gives me advice on anything, I really want to see what they have done before I even listen to them.

So, it's it's like everyone is giving the advice nowadays on something that they have read in a book that they haven't really practically done. So, yeah, I I really love your answer there. I think it's refreshing, so thank you for that.

Oh, sure. You're welcome. But it's it's it's right, you know, you you see some guys who, even if they did great things in development, certain software, or got a really big bailout or got a lot of money for their plant, they give you advice for renewable, you know, for all kind, you know, how to make energy, what is efficient, and it's like I'm wondering, "Wow, what kind of knowledge they have to have?"

Well, I wouldn't, you know, that's why advising it's yeah. It's I don't Anyway, I I gave an answer on that already, but you get me what I'm saying.

It doesn't lead to innovation. Advice only leads to repetition of the past, and, you know, if we want to innovate, we have to create, and that that doesn't happen from the past. I think that comes as a spark, and you just have to keep working and finding new things, and full stop.

Yeah. But, but even to be honest, I like to listen to advice. You know, I like to listen to opinions because sometimes it helps me to develop my own opinion. Doesn't need to be the other opinion. So, actually I like to listen to all kind of different opinions, and that's helping me finding my own way. But this is only how I going to use it.

That's a good point. I think I I'll just rephrase what I said. So, if you have a challenge, then you go out and see what people have done there. But before you have a challenge, you don't fill yourself up with knowledge that's going to just make your life difficult in terms of choices you make. So, okay, with that, I'll end my Okay, I'll rest my case before I get it wrong. Okay. I have to ask you one other thing. So, you said earlier that, you know, from tech perspective, the biggest challenge was to find people who would, you know, take your monolith software or the back end to event-driven architecture. So, could you could you tell me a little bit more about what led to the need for the event-driven architecture? Was it the computing needs, or the the amount of, you know, system breakdowns you were facing, or you just decided, "Okay, with the growing customer base, we are going to have these pains, and let's change"? Did you do it ahead of time, or did you wait for problems to happen before you decided?

Two things. First is, we saw already we got some problems. The good thing is with solar, those data are coming always on a same data data level. So, it's actually at that point not that challenging. It starts to get challenging when the units, we call them gateway, which are sending, pumping up the data, somehow and they are storing the data. So, if the connection is lost, they still been stored, nothing is lost. That's nice redundancy.

Then you get a big chunk of data all of a sudden.

Bingo. And then you all figure out, "Oh!" And this is where your system got challenged, and then we had a couple of times where we saw, "Whoa, this is not what it should be." So, on the one thing is really to get those data in the system when things got come, you know, got a lot together. That means to scale there quickly was not really available.

And then and the second thing is, when you're just putting writing those files away is all good, those data, but then do analytics on it. And you want to have analytics even quickly. I mean, I mean, with analytics is not, "Yeah, we do a certain machine learning." I mean, how is it? I mean, you have an idea, okay? So, you're saying you pull some data, you want to see, "Well, by the way, what is voltage doing on that temperature? Is there a" So, you first need to dig in, you need to smell it, you need to see what's happening. You need to pull another plant, and it that needs to be fast. If this process is not fast, then, you know, your idea is lost. I mean, if if this takes a whole week, I mean, how many things you can check?

So, but if you have something quickly, it's go through, give me this data, that, "Oh, that's interesting. Oh, is it there?" And if you have it fast available, you can make really quick assumptions. And then you can work out You see something, and then you work on some machine learning or AI stuff to dig in deeper. But you quickly need to, you know, smell it, check it out. And that requires fast systems. And and that requires a different way how to deal with data.

I see. So it was very much a you had a couple of problems, but mainly the use cases kind of dictated the the speed which you had to benchmark, and then, you know, what's the best way to do it. And here it was. Okay. Great. Okay, so I think we've come to the end of the episode. I hope you really enjoyed it. But I I really enjoyed talking to you today. I think it was it was great. Like just like the first time, by the way, so.

Thank you, Luke.

Thank you, Ingmar.

Xenia, yes.

A special thanks to our podcast guest this week, and once again to our sponsor, Sphere Partners, for bringing this episode to life. If you enjoyed this episode, drop SphereCast a five-star review on iTunes, and share this content with your network. For any relevant links or notes from this episode, check out our podcast website at www.sphereinc.com/spherecast. And always remember: when you think you can't, technology can. See you next time.