Welcome to the Nordic Hardtech Podcast. The show where we explore the people, companies and bold ideas shaping the future of physical technology in the Nordics. We speak with founders, engineers, investors and enablers building the Hardtech ecosystem. Read more and subscribe at Nordichardtech.com. This episode is recorded in collaboration with Lightbringer. Protecting your IP is crucial but often expensive and complex. Lightbringer offers an AI-powered platform simplifying the patent process. Nordic Hardtech listeners can now try it free for 30 days at Lightbringer.com with code Nordichardtech. Very much welcome to Massa Becline, CTO and co-founder at Flux to the podcast. Thanks for having me. Happy to be here. Flux, you're building AI-powered systems that left humans and wildlife co-exist. Can you just mention in summary what does that actually mean for the people who don't know the company from before? Yeah, we like to start by saying that we actually speak to animals using a language that they understand and how we do that is leverage technology and like you mentioned we use AI to infer behavior of the animal and also what type of animal it is, what species it is and then based on that information we play back a acoustic signal that they understand and then by doing that we can move them away from areas where they're not supposed to be. Like infrastructure, good example would be roads, rain roads, airports etc. Very cool. We'll get into some of the details soon. So you have a very interesting background. You spent about 25 years in California, right? How did you end up there to begin with? Yeah, it was a long time ago but I was playing American football here in Sweden and which seems odd but there is such a thing and I've got someone told me about it. I think it was an American football coach that was here coaching us that told me you know you could play football and go to school in the United States and that was about the time I finished school in Sweden so I decided to take a chance at it and I ended up in Los Angeles. I was into computers back then already and I want to be close to where things were happening, Silicon Valley etc. And yeah that's how it started. And you spent about 25 years, what did you do, what kind of jobs did you have that type of companies did you work with? Yes, my back was in software and I started out as building applications on the web. Back in the late 90s everybody wanted a website but they're also one of the websites that was easy to manage and this is before WordPress and other types of tools so you had to build everything from scratch. So that's where I started. I quickly moved on more to platforms and I joined Disney and worked on a content delivery platform for them and that led to a job as one of the first engineers at MySpace who was just the first social network. Quick question, do you know Tom? Yes I do. Yeah I actually worked very close with him. In the beginning there was only maybe like 25 of us. Oh amazing. And he was the one running everything basically not the business side but anything with product. And it was very basic as far as I was responsible for their music homepage and I was talking to him daily about okay what's the top 10 list of songs and you know who she would be featuring on this page way before you know more intelligent systems that automate all that stuff. Yeah. So amazing. So you spend a lot of years in the US and then you move back to Sweden about four years ago. Something so how did you come to join Flux? Yeah so moved back to Sweden and didn't really have a network here in the tech industry so I was attracted to some incubator I figured I would be a good way in and I found the company that was part of Sting and didn't stay with that company for that long but during that time Flux was part of Sting as well and I met Sarah my co-founder there and we started talking and after a year or so we decided to join up and do it together. Cool. And you mentioned being in the intro around the kind of core innovation and the technology that Flux actually does. How did that technology come about? Was it a kind of research driven process? What's something you started testing? It was based on it's always been based on research but the idea came about Sarah had this idea where drones were being used for all kinds of stuff or trying to be used for pizza delivery and stuff and she worked I believe was she's the science city at the time and was involved in a lot of drone projects and she felt like why can't we use it for something good and at that time wild boars were a big deal in Sweden it was ruining crops for all the farmers and they started to join up with two other guys from KTH working with AI and robotics and they tried to use drones to keep wild boars away from crops that's how it started and in the beginning they actually tried to build their own drone. So it kind of started as a drone project but then went more into acoustics? Yes the acoustics was there from the beginning. Okay, so they did some research and found proof that you could actually use acoustic signals to communicate with animals and make them understand like move away or come here for example right you can use their warning calls to keep them away or there's other calls you can use to attract them so the theory was that why can't we use that put a speaker on a drone and then kind of heard them to an area that's safe. And for someone who's not into kind of animal sounds a lot like what do you actually do? Do you like play back actual animal sounds to communicate with them or is it some kind of AI process around that? So it's both. We will do leverage real sounds but as you can imagine the actual warning calls and stuff it's very sparse and it's very hard to go out and capture. Yeah, so we are actually doing our own sounds now with leveraging general AI as much as possible and also just fine tuning existing sounds and changing a little bit and then we have the best one with that's our edge product is great for us to also to test all these sounds. Yeah, and how effective is that acoustic process and actually like media animals or making them come closer is there a number on that? Yeah, so we know the edge system that we've been testing now for two years are prototypes. We have a very high success rate. It's like 95%. Okay, and the combination there is using the right sound but it's also the surprise for the animal because you have this box sitting in nature right and it's quiet. There's no smell. There's no other signals for the animal to pick up on but then you play the right sound and it's a startling and it's also a sound that they understand what to respond to. Okay, a problem with a lot of these systems is that they use sounds that are very abrupt or very mechanical and this animal sometimes the reaction is that they freeze because they don't know how to respond to it. So what we do we leverage their authentic sounds and that enables the animal to make a decision quicker. Yeah, it sounds like a very noisy kind of development environment that your office is there a lot of animal sounds on time or could be at times. Yes, I think usually the person working on it has headphones on. But sometimes we do test it but we also have a lot of units deployed so we prefer testing it outside. Yeah, and it's been calibrated for like the most common animals in Sweden or is it like generic for all type of animals? No, so it's very specific. Right. But we're focused primarily on Northern Europe and in Northern United States. Yeah, so all across the tiger. Exactly. The animals are most sweet, very familiar. Okay, but we've tried other animals too and we're working with customers all over the world right now. But our focus is really, yeah. Cool. So that's a bit about the core technology then like how are you actually applying this technology into products? So you mentioned drone space for example, you mentioned an edge product. Yeah, what are you? How are you? Yeah, so the drone space we have actually parked for now based primarily business reasons. It's a hard time for us to make inways with the right customers. There's a lot of regulation. Yeah, anybody who's worked in the drone space will understand. But the edge product is we see it almost like our beach head product. It's very easy for us to deploy. It's cheaper. It's now battery-driven. It can operate autonomously and customer will see results right away. So that's we're seeing great success and a lot of customers want it. And that edge product that's basically a little box that you can put up on a fence or on a pole or on a tree or whatever it is. Yeah, but yeah, it's not big. It's pretty light.
and very is to mount it usually around two meters height, just to make it optimal with motion detection sensors in the camera. And it operates in any environment. I mean, we tested it up in Kirinna. So, you know, we would try to go down to like negative 40, at least. And we even, we picked a battery now that could be charged in very low temperatures as well. So, we're optimized for Swedish and North American, you know, cold climate. - Climates. - So, you are, I mean, you have a little box, you have a battery, you have a, I guess, a 4G connection or something like that. - Yeah. - And then you're pushing that data up to a cloud, I suppose. - Yeah, yeah. Yeah, so one thing we learned from our prototype, which we have had, you know, in two years now, is that 4G is great, but not available everywhere. So, we're actually now using an LTEM technology, which enables us to get a lot more reach in areas where, yeah, 4G might not be available. But, and we also learned that this possibility that the device, the box, might not be able to connect up as well. So, it can operate completely autonomous and in our way, the onboard AI. And then when it's connected, it can then push up the data. So, you can see reports and analysis and stuff. - Yeah, ideally, we do want to see the data, right? But we made it so it doesn't have to push it up. - Okay. - Yeah. - Is there like a ledeliclation problems with this having cameras and microphones all over or is that something you're getting around? - Yeah, I mean, we have to make sure we, you know, GDPR, we fall within that. So, any human pictures, we actually don't push up into the cloud, we just destroy them on device. So, they never actually, basically never exist. Besides in memory for a short period of time. So, that's one constraint. Otherwise, yeah, our customers probably have to pay attention to where they put them up. You wouldn't be able to put them up in the middle of Stockholm, probably. But usually they control the areas and it's usually their, you know, their infrastructure. - Yeah. So, this edge product, was that something you had like an early hypothesis on or something that you have developed to get with the customers. Tell me a bit about that process. - Yeah. I think, I mean, the need came from customers and traffic packet. They needed help by keeping a reindeer out of from the roads up north. It's a big problem because you know, you have the wildlife fencing, but they always end somewhere. And that's where the problems exist. 'Cause what happens is they walk around and then they walk inside of the wildlife fence and yeah, it becomes a very stressful situation for the animal and a lot of accidents happen. So, they wanted a solution that could actually sit on the edge on the wildlife fence and keep them from entering that area. - Okay. - So, that's kind of where that idea came and it was a quick prototype process. I think it took around three months to build it, three-deprinted, obviously outsourced the actual PCB, building those, but everything else was built in-house and mechanically put together and shipped to our customers. - Cool. And it was the process from that inception to actually having the now kind of launch product that you're up to have now. - So, yeah, so that's been a longer process. Partly because we took a while for us to get all the learnings from the first prototype, but once we started to understand, we need, we knew we needed to have an excellent battery life so that started the process of looking at what type of components and what we need in order to actually operate for, and we set goal of like minimum six months, operation time. So, that's put a lot of constraints on the system. And then also we knew based on this prototype that we need to get more range, 'cause the motion and the sensor we used on the prototype had a range of best case 15 meters. - Okay. - And you can imagine if you need to cover an area that means you need a lot of these units. So, we set a goal to have 40 meters and we knew we couldn't do it with off-the-shelf stuff, so we have to innovate on the lens design, on these motion detectors, but also leverage multiple motion detectors, and we have filed a patent on that technology, which is enabled to get three times the reach of anything else out there on the market. - Wow, very cool. So, in this whole process, let's say, with the edge product, what has been the most challenging part from a technical perspective on that? - I mean, obviously the motion detection design took some time for sure. Also finding the right hardware for the onboard compute AI. We're using microcontrollers because we need to be able to sleep 24/7, and we can have a Linux-based board, 'cause boot up time is too slow. - Okay. - 'Cause when we have an interaction, we need to wake up the system, and we can't wait for Linux to boot. So, we knew we had to microcontrollers, and to be honest, that having compute on a microcontroller is pretty new, but we managed to find something after some research that actually could work, but that was a challenge. We picked some, that first microcontroller we picked, we finally had to give up on, 'cause the manufacturer ended support. So, we recently had to switch to another microcontroller, which it causes a lot of headaches, especially on the software side. So, I would say that's been a challenge, operating within these constraints of just needing to have, use microcontrollers. - Yeah, got it. Let's look at the business side of this. So, you mentioned some of the customers that you have for these systems, but like who in general is the ideal customer for you? - Yeah, I think we broadly will look at as infrastructure, and we have a lot of momentum on airports, and they have very, very concrete needs, a lot of safety and costs associated with animals on airports, so you can imagine, but also roads, like I said, we work with traffic that get in Sweden, and we even start to work with similar businesses in the US, and then, yeah, rail roads as well. So, that's our infrastructure right now. Long-term, we actually do see agriculture as the biggest customer by far, if you look at market sites, but we believe we need to probably innovate a little bit more, and even work more on our price, and also get back into the drone space, to actually really serve the agriculture. - Cover larger areas, etc. So, speaking of price, how you thought around the business model and the pricing for these products? - Yeah, so we've basically been doing value-based pricing. Sure, we have some type of price target, but customers have so different needs, and also the value is so different, if you compare an airport to maybe some other smaller type of infrastructure. So, that's how we've been looking at it, and we're selling it more as a, we have a platform, we call it the Wildlife Platform, where they can actually access all this information, and that's what we can plug in our own edge product, but also other types of hardware products in the future, our own and potentially others as well. We're looking into potentially partnering up with other wildlife cameras and stuff. And so, we're selling that as a software, as a service basically, of course, with a yearly ARR, then. - Yeah, nice. Has that been a problem finding those right price levels with customers, you mentioned that they differ a lot between different segments? - Yeah, I think, yeah, I guess it's always hard, 'cause we're also kind of in a new market. Sure, they have used a lot of scarecrow type of technologies, but they operate very differently, very few of them, are looking at this from a platform perspective. It's more like here's a piece of hardware, put it up, and hopefully it won't work. So yeah, it's been a little bit, a little bit of a searching there to figure out the price, but I feel like we're getting better at it. - Good. Good. And how are you then actually selling this? Is it like going directly to these airports and train companies, or are you using resellers or agencies somehow? - There's a lot that direct B2B. We are now talking with some other types of partners that can potentially help us. Airports as well, we've worked with Svedavia, for example, and there are other types of Svedavia's in Europe, for example, in New York. - Yeah, it's just one in each country. - Yeah, yeah, yeah, yeah. So those are ideal customers, right, 'cause then you reach more customers and more airports in that case. So we have a dual strategy there, I guess you could say, where we go direct, and then we also try to find.
and the partners that can help us. - Yep. - Do you have your own kind of sales team, or your own commercial team? - Yeah, we are. That's something we're actively building out. - Okay. - And we just made a hire in the US for someone that's gonna take charge of that. - And how are you looking at this from a geographical perspective with respect to the business? I mean, you started in Sweden, but you seem to do a lot of work in the US as well. - Yeah, so very early on, we got a lot of inbound from the US. So we've been working in the US for quite a while now. And my co-founders and the CEO Sarah is actually there right now. We are part of a deep tech incubator in New York called Genius New York. And we now incorporated in the US as well. And like I said, our CRO, we just hired, is from the US as well. So the market there is obviously huge. And they're a little bit more aggressive on the business side with new technology, quicker to adopt and more easily to get pilots going. - Okay. - So yeah, we found it very, it's an opportunistic move 'cause we see that the response is a little bit quicker there. And we feel like anything we learn there will easily adopt here as well. - Okay. And how did you like, when did you make the decision to actually go towards the US market as well? Was that a discussion or was it kind of clear that you have a lot of inbound? So let's go there. - I think it was very clear. Sarah's always been interested in that market as well. And obviously I spent a lot of time there. So it was very natural. It wasn't a long discussion, especially when we had to make the decision about this incubator. And part of that incubator was that one of us had to move. And yeah, it was easy for Sarah to make that call. - Yeah. - First to make it the call. - Makes sense. - Let's look a bit about the, on the funding journey of the company. So how you done that? You've taken in a number of different kind of grants, some angel investments, and some more strategic investments as well. Okay, can you tell me a bit about how you thought of that funding journey? - Yes, we started out with a small pre-seed round with Almi among others. And then since then we did a also small seed round early this year with unconventional ventures. And you also had an orchian who's been part of the journey. So we've been fortunate to have some good partners. Now we've seen, we have a lot of traction right now. So we saw an opportunity to actually extend the seed round. We call it the seed plus, which is something we're working on right now. And we probably will close here in the next week or so. - Okay, interesting. - So yeah, it's been, yeah, fundraising is always interesting. - Yeah. (laughs) And it takes a lot of time. - Yes, it does. - Have you, it sounds like you have been fairly successful in that space up until now. What do you think has been kind of the, what has been the prerequisites for being successful in funding this type of company? - You have to be persistent. You have to have a lot of patience. And you have to realize it's not a software, you know, ECSAS company, where you can easily just show some ARR and investors will be get excited. - Yeah. - Where they admit it or not, hardware is something investors look a lot harder at. - Yeah. - 'Cause they know and the risks associated with it, right? - Yeah. - But I feel we've been, we've definitely been lucky, but it's been mostly for a lot of hard work and sticking with it. And to be honest, more of a support role there. So Sarah has to want to have been banging her head against the wall. - Yeah. - A lot. - And do you have any kind of input on, do investors in Sweden versus Europe or the US see this type of company differently somehow? - Yeah, I mean, my experience with hardware is primarily Sweden, so I can't compare it on the hardware side, but on the software side. I would say, and it's definitely, it moves, tend to move faster in the US. There's also a lot more, it's just a bigger market when it comes to venture capital. - More capital, more capital. - Yeah, but that side is also a lot more competition, right? So, but in general, I would say it's, especially at this earlier stages, it moves a little bit faster in the US than in Sweden. - In Europe in general. - And with that money now, with that funding, as you mentioned, the seed plus round, for example, what are you actually investing in from that money? - A lot will go into just building the business on the business side, but also we are doing some critical hires on the R&D side, and that's something we do here in Sweden, so all R&D is in Stockholm. - Okay, nice. And how is that market from a talent perspective in Stockholm, do you think, from the type of products that you're building? - The talent is great. I mean, it's really good. I'm actually been surprised by how good it is. I really enjoy working with engineers we have, and the meeting engineers, potential engineers as well. And we are actually lucky enough, I think, that we're also in a company that's very interesting to a lot of engineers, you know, we work on the edge with the latest hardware, and we work with AI. We're really pushing the boundaries on a lot of things. - And there's also an interesting kind of sustainability and resilience perspective, which I think some people would find interesting. - Yeah, and most people like animals, and they see the conflict of animals, as it's growing and growing. So yeah, it's attracting talent that's been pretty easy for us, which is nice. - And you feel that there is also a hardware, a hard tech scene for talent in Stockholm, where it's that, yeah. - Yeah, especially, I mean, we're literally on campus at KTH, and I think that's probably one of the better schools, yeah, in Europe for sure. So yeah, we're very fortunate. - That's perfect. So, I mean, you've been around before, obviously, you've worked in like 25 years in the US with a lot of different startups and companies, even my space I know now, which is a good thing to know. What's something that you learned from the flocks you're in over the past couple of years that you didn't know from before? - I mean, yeah, everything with hardware, but that's no news to this part, I guess. Not to me, it was, I think the biggest takeaway is that in Sweden, and I can only speak for Sweden, I'm in Northern Europe, is people are a little bit more careful, a little bit more methodical when building, which is great, and I've definitely leaned into that, but I also been trying to push a little bit more towards, you know, let's just build something, test it, get it out there. So it's been a good experience to balance those two. - Do you think we need more of a build fast on the break things kind of mindset in Sweden as well? - Yeah, maybe not the, yeah, I'd love to build fast, I'm some big fan of the break things, but build fast and test things maybe. Yeah, I think the break thing has gone a bad rap and for a lot of reasons, good reasons. - Exactly. - But definitely build and test. The iteration part is key, and it's amazing how fast you can iterate now on the hardware side with three to printing and very quick turnaround on PCBs and stuff. So, yeah. - Exactly. So, what would you have any advice to founders or entrepreneurs trying to build similar types of companies out of Northern Europe, what do you think they should consider? - I think you just need to get to the prototype as quickly as possible. Before that, you really don't have anything and it's the same on software, you need to get something build the customer can test. Obviously, that's harder with hardware, but the tools are there now, and something we even discovered in the last couple of months now, even like the LLM. So, when it comes to coding for embedded. - Yeah, even for embedded. - Getting much better. - Okay. - And that we've seen a big shift just in the last couple of months. - Yeah, and is that kind of generic tools like Cloud or is that something something that's embedded? - The team is all over the place. Everybody has their favorite. - Yeah, sure. - I think so. On the embedded side, it's the model from OpenAI, the latest one. - Okay. - And that's been best. You know, it's hard to, I'm doing air quotes here about, you know, it's getting harder and harder to compare the models now, I think, because they're getting better. But there's been a shift, and we've noticed it, because we've been bowing against it, especially on the embedded side, because the results have not been that great. But the last couple of months, it's better. - That's really, really promising for our own or our founders as well. - Yeah. - Would respect to the kind of the funding, and if you have any advice there, like do you have, when do you think these type of companies should look for their first investment? - Even if you're not looking, you should probably start working on it right away. You need to make connections, you need to get out there, you need to talk about your product. Even if you don't have it, you need to get the word out and make that personal connection. And then get as quickly as possible to prototype and have some, even if you can't sell it.
get some good customer feedback that you can share with investors. Yeah, you need to, that's really quick, quick, you can get out of better. But on the rest of the social aspect. Thank you for that advice, Thomas, and thank you so much for joining the podcasts. It's been great. [Music] Thanks for listening to this episode, which was recorded in collaboration with Lightbringer. Don't forget, you can simplify your patent process with Lightbringer's AI-powered platform. Try it free for 30 days using code
[email protected]. Thanks for listening. Read more and subscribe at NordicHeartTech.com. [Music]
PodcastTranscript AI