Yeah, just imagine you're learning to drive and your instructor tells you, "The highway exit is coming up ahead; now, about 1 to 2 seconds before the exit, you have to turn the steering wheel about this much and this much, because then, 1 to 2 seconds later, it effectively makes a curve of about this much and this much." You wouldn't drive like that with your car, but a conventional variometer is exactly that. The flight instructor tells you that you're flying into the thermals, you feel it, it goes up, then the vario starts beeping, then you count 21, 22, 23, and then you turn. And it's pure luck whether you can center the thermals like that or not.
Podz-Glidz, the Lu-Glidz podcast.
Stories from the cosmos of paragliding.
Today with Conny Schafroth. And I'm Lucian Haas.
The most important instrument for a paraglider is their variometer. With beeps, it shows them whether they are currently climbing and where they should circle to gain valuable altitude. However, classic variometers don't exactly make it easy for pilots to read the air precisely. Because there is typically a technically caused time lag between an up and down movement that you can already feel with your physical senses—that is, your popometer—and the corresponding beeping of the variometer. In large, constant thermals, it doesn't interfere much, but when it comes to centering tight, turbulent cores, the delayed auditory impression of the air movement is counterproductive. The Swiss pilot Conny Schafroth has been bothered by this flaw throughout his flight career, which now spans more than 30 years.
In 2015, he suddenly had the solution to this problem in mind. Instead of just using one air pressure sensor to record flight altitude, he integrated a whole range of additional sensors into his Vario variant called XC-Tracer. Accelerometers, belt sensors, gyroscopes, magnetometers, GPS. He combines all that data using complex mathematical formulas to essentially capture every position and altitude change of the paraglider in real time. That way, the Vario and the popometer match up. In this episode 117 of Podz-Glidz, Conny Schafroth doesn't just give insights into developing and testing new Vario functions. I also talk to the 58-year-old about his earlier days as a paraglider designer, a special aircraft project called Smartfish, and why the paragliding scene would benefit from taking to the skies with less performance and more experience-oriented focus.
If you enjoy the podcast, please support my work as a patron. You can find out how to do that very easily on my Gleitschirm-Blog Lu-Glidz, specifically on the "Fördern" page.
Conny, most people probably know you as the developer of the XC-Tracer Varios. However, if I'm correctly informed, you also developed your own paraglider many, many years ago. Tell us the story of how that happened.
Flying has fascinated me since I was a little boy.
Before that, I built a lot of model planes, and then at 16, I learned soaring. And then I was at the university in Zurich, at the JTH, and suddenly there were paragliders. But the paragliders, those were more like... well, back then they were still parachutes. And I had the impression that you could do it better.
And back then, there was a paraglider manufacturer right nearby. So I contacted them and designed a paraglider. Then he sewed it together and I went out and flew it. So, in principle, I learned paragliding with a self-built paraglider.
Did you already know anything about aerodynamics back then? Did you learn that during your studies, or was it just some gut feeling? You have to learn that somehow. You have to be better at it somehow.
No, no. I obviously had aerodynamics in my studies. I was also in the wind tunnel and all that. So it was quite well-founded. The first ones flew pretty well, too. And my brother flew across the lake from Niesen to Oberhofen with one of the first prototypes. That was a sensation back then. Didn't land in the lake? No, he flew over the lake. It worked. What glide ratio did you need for that? Yes, I think it was six or something. What year are we talking about? I think it was 1988.
And that means, I mean, back then, those were still the early days where everyone was always squinting, saying performance has to keep getting better, and the glider that glided the best was the one that sold the most. Was the glider you designed back then actually a success?
Yes, it was pretty good. For that time, they were very good gliders. Looking back now, I have to say it was a total catastrophe compared to today's gliders. But for that time, they flew very well, yes. Under
under which brand name or what product names, I mean type names, did it hit the market? So, which glider can you associate with you, even if it doesn't say so anymore today? That was, back then, that was...
The brand was called NorthSails. And I believe it was later renamed to Rico Gregorini AG for some trademark reasons. And there was the C-series, the Solution—I don't know, it wasn't that long ago. I don't remember exactly.
But were you basically the designer for them? Or one of the designers for them?
No, I was the designer for them and also the test pilot at the same time. And that was pretty nerve-wracking because at the beginning, you didn't know anything at all. So we had this phenomenon called deep stall. Back then, nobody knew what deep stall was. There weren't any gliders either. We once bought one from the local hang gliding instructor. We bought a glider for the hang glider. And when we tried it out over the water for the first time, it immediately didn't work. It didn't open at all, and all that. Yeah, those were relatively wild times, and I was also pretty lucky that nothing happened.
So you're saying you survived, but you weren't unhurt back in those days either?
No, well, I did do one stupid thing once when I wanted to land right in front of the company. I overlooked a telephone cable. Then I had to make such a U-turn and it didn't go well. It wasn't enough. I ended up with some pretty bad back pain. But other than that, I've actually never had anything.
How did you design paragliders back then? At that time, there were no CAD programs or the classic paraglider programs—meaning design programs like Gliderplan—that everyone uses today. And where there are actually industrial standards that very, very many people work with. What did you have back then? Was it all hand drawings or how did you...
how did you do that? No, no, it wasn't hand-drawing, because I was lucky enough to have just finished my studies at ETH. And they had a drawing room there that I still had access to. And they also had this huge plotter. And with one of the first CAD programs, I was able to draw the gliders. But it was nothing more than just profiles and then laying the paths over them. And you had to manually correct everything by hand, calculate everything by hand, the lines and so on. Yeah, it was a pretty restrictive process.
Why didn't you continue with that career? I mean, why didn't you stay a paraglider designer? You might be one of the most famous paraglider designers today if you were already performing at that level back then. Yes,
That wasn't really for me. It also involves test flying and all that. I thought, no, not at some point, that's not so great.
Maybe also because designing paragliders isn't exactly a precise science?
No, that was actually still fun. It helps me to this day because back then, you just didn't know anything. Mostly, well, there were no GoPros, for example. You couldn't record anything. If you were lucky, you had a very grainy video recording from boats of your test flights. But a glider will increase your flow. And if there's then a flow separation because the angle of attack is really too large and the flow breaks away, and because of that, the pressure in the glider is no longer sufficient to maintain the profile, or if the pressure is no longer sufficient and because of that the profile gets deformed and because of that there's a flow separation. Those were questions; we didn't know that back then. You had to figure all of that out, but it was actually fun in a way too.
Is there anything regarding the actual paraglider, among the things you designed, where you were somehow in the bronze category?
No, I wouldn't say so. I think Fallhacker also made the openings on the underside at the same time as we did and so on. But no, I wouldn't say it was industry-defining. What did you...
what were you looking for for yourself back then when it came to flying? What was it that interested you so much?
Good question. I've actually always enjoyed flying. The funny thing is, I'm not vertigo-free. So, something like climbing up the Bernese Oberland to the monastery, that's not really my thing. Climbing didn't work either. But flying, that has always just fascinated me.
And you don't get dizzy? So while flying itself, you're sitting in the harness and you always feel comfortable?
Yeah, not always. I'd say about one minute per hour, I feel it's maybe a bit too turbulent. I don't like it either when it gets too high. But there's a story about that, it shaped me. We were, yeah, that was, it was somehow the late 80s, we were at Niesen. With a southwest wind, good conditions actually, wonderful. You mentioned a few laminals. And then you could fly almost the Düden along the slope, it was really, so high soaring. And then fly out and then spiral down and then high soar again. Then I flew out and suddenly the glider was gone. And then it really, it just came out of nowhere, which normally in the first year, oh now it's turbulent, now this is coming, now you might have to grab the brakes. And that just completely wrecked the glider for me.
The glider was oriented low. It was just so slanted in front of me. That's what caused it to slam down so diagonally. And my first thought, looking down, was, oh, throw the reserve parachute now, that would be really bad above the Niesen, above the Kette. Strangely enough, I wasn't even scared at that moment. And then two seconds later, the thing was open again. And that's when the shock hit me. But in the first moment, it wasn't bad at all. And then later, they reported on the news that day that there had been a strange phenomenon, that the jet stream had reached very low down.
So you basically hit the jet stream turbulence? Yes,
exactly into those last bits of turbulence, I have no idea if it was still the jetstream, but it clearly caused a very turbulent atmosphere. And that was pretty special, when you just suddenly fall into empty space and don't notice anything beforehand at all.
Did that change anything about the way you approached flying afterwards?
Since then, I haven't enjoyed it as much when it flies high. I mean, I don't like flying that high over the terrain as much. Because you think, maybe the jet stream will get too close again? No, no, no, I don't think about the jet stream at all. No, it's just, I don't know, maybe something has remained habitual. Just being that high above the ground, it's not
...as good. But I mean, you've been flying for over 30 years now and you're still flying. What appeals to you about it today?
It's a very, how should I put it, a very emotional matter. You see landscapes you'd never otherwise see. Yes. You meet people you'd never otherwise meet. You have encounters you'd never have. It's almost like a doorway into another world, I sometimes get the impression. It's hard to describe. And what kind of flyer are you? How would you describe yourself? I'm not looking for the mileage, that's not really my thing. I'm looking for the beauty, so to speak. So I enjoy being able to go to Ohran at midnight somehow. At midnight? During a full moon or what?
Yes, full moon.
Exactly, that's cool, it's fun. Exactly, for example in the snow. That's incredibly beautiful. Sometimes I also go in the dunes. I really like that too. Or on Sundays, I fly over the glaciers. I really like doing that too. There are already many different aspects. The main thing is that it's emotional. You've never gone on a long distance? Yes, you have. I've also flown 100 kilometers over 30 years. But I'm not really interested in that. That's not really my thing. 30 years ago? Yes, yes.
With those 21 cells they had back then or whatever. Exactly,
yeah. How
Can you manage 100 kilometers there? Was that also a day where the jet stream was so low that it pushed you forward properly, or what? No,
No, no. It had nothing to do with jetstreams. No, I just started in Viersch and then headed up from Goms. That's the usual race route, up from Goms and then down in Graubünden. Until just before Chur. That was about 100 kilometers. Yeah, there was a bit of a west wind, but it actually went well. But like I said, it's not really my thing. And then driving back through half of Switzerland until evening, until you're back home and all that. That's not really... yeah.
That means you probably won't ever find any flights of yours on XContest?
No, you'll never find anything there. No.
I find that interesting because how does someone who doesn't really care about performance in the end—at least it seems that way now—how does someone like you, who isn't looking for performance at this point, come up with the idea to develop a particularly high-performance vario?
I was always annoyed that variometers have such a time delay. And for a long time, I tried to convince a buddy who worked at a startup to build something like that. But somehow, I couldn't really convince him. And then, what was it, about nine years ago or so, the weather was extremely bad at Easter and I really wanted to go skiing and stuff. And I thought, yeah, you'd go crazy if you couldn't do anything. So I bought a microcontroller starter kit as a way to pass the time and started programming; I got a first LED to blink and was really proud that it worked. Then I got a few LEDs to blink, showed them to Lisa and said, yeah, look, cool.
And then she asks, "What can you use that for?" Zero enthusiasm. I, I have no idea, I don't know. Lisa is your partner, just for clarification. Yeah, exactly. Then on Easter Monday, there was suddenly this little window of weather where we could go skiing, we went powder skiing. And then somehow, I don't know, two or three times on the ski lift, I suddenly say—I hadn't thought about it—I say, "Ah, now I know what I can do with it."
And yeah, it didn't take long; then I bought the first pressure sensor, then I bought the accelerometer, I experimented around, and then pretty quickly the second or third prototype was already significantly better than what you could buy on the market. That's how it came about.
That was like a "eureka" moment you had at the ski lift. But the thing is, the special feature of your Varius is that you don't just use a pressure sensor, but also something else. Can you explain that right now? But how do you come up with the idea of saying, "Ha, I'll combine these, that's the solution"? Does that really just fall from the sky?
into your head? I don't really know, somehow. I mean, it was clear that the pressure sensor alone wasn't enough, that there's always a time delay, and that you have to combine it with other sensors. Yeah, in principle, that was clear. But don't ask me why. I just knew you had to do it that way.
Before you explain a bit more about how the XC Tracer Varius works, or rather that specific feature you have, let's talk about normal Varius first. Because I think many people use Varius, but they probably don't really understand how they work because they don't pay attention to it; they just turn it on, it beeps, and wonderful, you can use it for something. So explain to me, what is the basic function of a conventional Varius? How does it produce its beep?
So, in principle, a variometer is an instrument that shows you your climb and descent. To do that, you have a pressure cell or a pressure gauge, if you like. From this pressure gauge, you measure the altitude over time. And if the altitude increases, you get a curve that goes up. Then you can say, okay, in the last second I climbed three meters, so I had a climb of three meters per second. And one point is, if you do it that way, a problem with a normal variometer is that this pressure cell doesn't always show exactly 1027.32 meters, but it's always plus 10, 20, 30 centimeters. It has what's called noise on it. And in order to filter that out, you have to use a filter.
At its most primitive, you could say you just do a moving average over the last ten measurements, and then you take that value and compare it to the value... the average value from before? Yes, exactly, from before. And then, if it goes up and beeps, you have to make beeping sounds, and if it goes down, you do nothing or you make a sink alarm. That's the primitive explanation. The problem is, every time you build in a filter, you have a time delay. And back then, I had this old vario from Bröning, like a yellow thing, like a shampoo bottle. And the delay there was about two to three seconds. So I just knew that when the thing beeped, I had to turn immediately. And that's what ended up being my downfall during the first test flights with the XZ Tracer.
So as soon as the XZ Tracer beeped, I knew, now you have to turn. Then I fell out of the thermal backwards. Until I learned, no, you have to wait until it shows less climb and only then turn. That was the turning, because I flew around with the old thing for so long. It was relatively difficult to get used to, but now it's no longer an issue.
So, classic variometers have a certain time lag. Simply because you have to calculate an average value from all the measurements, and that takes time, and then from an output perspective, it takes more time to get it filtered nicely. And then, basically, I only see what's actually happening about two seconds later.
No, the thing about the average value is a primitive way of looking at it. Of course, there are much more sophisticated or better filters than that. But still, you basically always have a time delay. You can do whatever you want about that, and it can be more or less significant. And that's the problem. Yeah, imagine you're learning to drive and your driving instructor tells you, there's a highway exit up ahead. Now, about 1 to 2 seconds before the highway exit, you have to turn the steering wheel roughly this much and this much, because then 1 to 2 seconds later it effectively makes a curve of roughly this much and this much. You wouldn't drive with his car. But a conventional variometer is exactly that. The flight instructor tells you that you're flying into the thermals, you feel it, it goes up, then the variometer starts beeping, then you count 21,
22, 23, and then you turn in. And it's pure luck whether you can center the thermals directly or not. But the XC-Tracers are very direct. That means you can just really wait until it shows a few strengths. Then you know you've passed the culmination point, and now you can turn in.
What did you do differently with the XC-Tracers so that you no longer have that time delay that's in the classic variometers?
have? There are different sensors built in. So it doesn't just have one pressure sensor, but it also has a three-axis accelerometer. It has a three-axis gyroscope. It has a three-axis magnetometer. And then the GPS data is also used. And all of this data is combined. And now, you can basically say in very simple terms, the accelerometer notices when you start accelerating, the pressure sensor shows you the average incline. And when it stops climbing again, it's the accelerometer's turn. That's a very primitive way of looking at it. Of course, it's not really like that. It's a very complex mathematical process. It took us a very long time to figure out how we had to do it.
But as I said, it's simply sensor data fusion and it takes a lot of work. And once you've invested that much work, suddenly it works.
You also mentioned magnetic sensor data in three dimensions and stuff like that. How does something like that come into play, for example?
Basically, I mean, each sensor on its own is relatively inaccurate. But when you combine all the data together, you get a pretty accurate idea of what your current flight path looks like at any given moment. The thing is, you don't have a single measurement, you have an estimate. And this estimate is based on all the data you can collect. And that's where the magnetic field sensor plays a role too.
It just measures the Earth's magnetic field and how you move within it, or something like that.
Exactly, right. And you need that to know your orientation, which direction you're flying exactly, and so on.
That means your vario actually knows, oh, it's flying north or south, west, east, and so on? Exactly, yeah. If you have a classic vario, where you say there's only a pressure sensor inside and it measures—couldn't you actually achieve perfectly sufficient accuracy if you just said, I'll take a pressure sensor that I can read 100 times per second, and then I have enough measurements in a short time, then I'd have a clean vario value after a tenth of a second, roughly. Why do I have to do something so complicated where I have to calculate three other sensors from three dimensions with all of that?
Because it's fundamentally better. So, if you only use one pressure sensor, you have the problem of time delay. You can do whatever you want, that's one thing. And the other thing is, of course, we don't just have the climb. For example, we can also calculate the wind very well because we have this flight path estimation in real time. You can do tech, for example. So you can do a lot of nice things once you have that. And that's much better than just a very simple variometer with a pressure sensor.
How did you actually know how to do that with this data fusion back then? Or is it really learning by doing and a lot of trial and error? Or how do you get there? I mean, you said at the beginning, yeah, I ordered an Arduino processor and got an LED to blink. That's probably a relatively easy task. And now to say the same thing—getting this microcontroller to calculate everything at the same time and then finding a formula where it provides me with a quasi-instant variometer function.
Yes, that was pretty complicated. Well, it's still complicated. Nowadays, we work with specialists to get it added in really well. I can't do that alone anymore. It's way too complicated. It's truly a highly complex mathematical task. It requires a lot of testing. You need to have very good hardware and you really have to go out and do a lot of testing.
But testing also means testing in flight, right? Yes, yes, flying. Yes, yes, flying. That means you're now, like you used to be a test pilot for your paragliders, you're now a test pilot for your varios? Exactly, something like that. And how can you test a vario in flight? I mean, you can say, okay, I feel it in my butt, it's going up there. But how do you know exactly that your 3D data fusion is outputting the data on how the air is actually rising and so, instantly? You can't simultaneously measure the air with another instrument very precisely to say, I have a reference now that I can then adjust my vario accordingly to.
can. No, you can't. You're right. That's relatively difficult to determine. That's where things like Sorten help, for example. Because there you can fly certain maneuvers, like a rocking motion, for instance. And the vario is supposed to stop beeping exactly at the moment you hit the culmination point, and you can see on the slope next to you that at that moment you practically come to a standstill, you're not moving anymore. That's a point of reference. And you can also take several devices with you and see if they all show the same thing and so on. It's not an absolutely precise science, that's true, but you can get pretty close.
Are there things where you can use a device like that, yeah, like on one of those—what are they called—the centrifuges, where you test out the G-forces? Ah, like a simulator? To say, ah, it measures exactly the right G-forces or something similar. That should then probably also be in your data sensors accordingly, the acceleration associated with it and so on. And that can probably be calculated relatively accurately then. So, from the centrifuge, you could maybe say the data is correct and I can then compare that with my vario accordingly.
That's how it is. But usually, when you're spiraling, you're going down and the vario isn't that important anymore. Yeah, you could go to a centrifuge, but that doesn't really help us. It's more about, how should I put it, it has to work in turbulent conditions, it has to work in calm conditions, it shouldn't oscillate, for example, it has to be stable, it has to run on different hardware. That's the kind of stuff we're testing. How is it, for example,
If I initiate a steep spiral? You get strong G-forces all at once, but then I descend, but with high acceleration and all that. Can an XCTracer detect that cleanly, saying "that's a steep spiral now," or will it sometimes provide me with slightly faulty data?
The first generation, yeah, it didn't always provide correct data. The current XCTracers, if they beep, you can assume it's really going up. So they can handle all of that.
But a steep spiral would be descending. So, can they handle the descent just as well?
Yeah, the funny thing is, they even get the wind calculation right. I once flew down from the Niesen, tried something out, and then I had too much altitude, did a wingover, and messed up a bit. And then when I was finished, it showed a completely different wind direction. I was pretty stunned, thought it was total nonsense. But it was actually spot on. So even when you're flying maneuvers, it can calculate the wind.
But maneuvers mean, I mean, normal wind calculation from Varius usually works like this: they say, okay, you have to fly circles, and then we—not shift, we basically measure the corresponding wind drift from the GPS data and can then say, okay, the climb was this strong, the drift was this much, and from that, you calculate the wind as basically vector math. How do you do that with your Varius?
I couldn't tell you the details. I couldn't talk about what's in the toolbox. But with us, it's not just circles either. You can just fly attentionally, that's enough. And that's because we have a magnetometer, so we know exactly how the little device is oriented. And unfortunately, I couldn't go into the details because wind calculation is one of our strengths. And it doesn't just work while circling.
Even if you don't want to go into the exact details right now. But you'll recognize things like circle radii and other stuff, and all of that flows into this calculation where you say, "Ah, he's reaching further out in this direction during the turn, and less far out of the circle in the other direction, so there must be corresponding crosswind or something."
In very primitive terms, yes.
Then tell me something else. You've also continued to develop the XC-Tracer Varius over the years and added new functions to it. You already mentioned wind with this Sohrenwind. If that's coming up right now, it's storming here. It might rumble a bit on the recording. Those listening shouldn't be surprised. There's a thunderstorm going through Bonn right now where I'm sitting. So, you have the XC-Tracer. The Varius has been constantly developed, and lately, at least the XC-Tracer Max 2 masters a function called TEC.
What's the deal with that? TEC, yeah, it's a long story. We've been working on it for years until we got it reliable. Imagine you're out in a glider. You're flying straight at, I don't know, 150, 200 kilometers per hour and now you pull on the stick. What happens then? The glider climbs and you're basically converting speed into altitude. And then a conventional vario says, okay, it's going up, now I have to beep. But a total energy compensated vario doesn't beep. A total energy compensated vario basically calculates the kinetic energy and the potential energy—so the energy from the speed and the energy from the altitude. And only if the total energy increases does it beep.
And in principle, we did the same thing with the Max 2. And the advantage of TEC is now, if you're doing a valley crossing and flying across the valley with a lot of speed and full throttle, and now you're—yeah, valley crossing is always a bit of a problem, whether you arrive high enough. And if you now fly into a small tube, yeah, you can't turn directly. You first have to get out of the speed. And then the speed is converted into height. And then it beeps anyway. But with the Max 2, it's now such that all of that is calculated out, that it really only shows when it's rising. Through the thermals. The rising through the conversion of speed into height, all of that is calculated out. That's why you can, in principle, center a thermal directly, even if you're going full throttle on the speed bar.
But how does the vario know that I just kicked the speed bar?
have? It doesn't know that. It just sees that you're converting speed into altitude. Then it reduces the speed. And the altitude increases. And because of that, it can calculate, yeah, okay, that's not really climbing. That's just a conversion of energy. I don't have to complain now. I don't know, on our YouTube channel you can see a video where we play around with tech like that. And there you see, you can rock the glider and let it lunge forward and rock it. And it practically doesn't move at all.
Does it actually make a difference whether I'm flying an A-glider or a D-glider? Because the more powerful gliders are usually much better at converting speed into height due to the profile and so on.
Yeah, for an A-glider, it doesn't make that much of a difference. The better the glider is, the more it matters. What we've found is that Tech works perfectly when you're flying straight ahead. And if you were to fly around with the thermals using it, you can try it out, you can adjust it. But it doesn't work that well because if the glider accelerates a bit, flies a bit faster but at the same altitude, that's enough. Oh, the total energy has increased, so you're climbing. Now I have to beep. And then sometimes the vario beeps before you even start to climb. That's not really intuitive. And that's why we've done it so that, in principle, you can set it so that it has Tech in straight flight, that it's normal X-axis in the thermals, and there it has a transition phase,
where it goes from 100% Tech to 0% Tech in a smooth transition.
But the pilot has no influence on that during the flight itself; he just knows, okay, if I'm flying a straight line, as you said earlier, a cross-country section or something, and then I enter a climb and let go of my speed bar and I'm still climbing, the X-adresser with the Max 2 and the tech function recognizes that accordingly at that moment, but after about three seconds, it's switched off.
Yeah, because you've also lost your speed by then, and after three seconds it's gone, so it's actually good if you have the completely conventional variometer again. Just a simple X-adresser, without the time delay.
Do you really need something like that? Whether it's three seconds earlier or later, I mean, I know if I'm properly in the thermals or something. If you're already saying, "Yeah, fly into the thermals and count to 23 or something," that's about the same amount of time anyway.
No, the thing is, if you're directly—you can basically center the thermals directly out of the speed bar. Not by luck, but you actually know, you can turn in deliberately. And you can't do that otherwise because you simply lack the information. Maybe you can compensate for it with experience, or partially compensate. But if you really display the radio where or how it's currently rising, then you have a great chance of being able to center the core directly. But I'll have to ask the top pilots who fly with them. They basically say, yeah, it just helps to map the thermals better. It's easier.
But then it's probably mainly interesting if I'm actually a competition pilot. Like, when I say, every three seconds I'm somehow really in the thermals and already ten meters higher than the others in the same core, and I'm sitting on top of them—do I just have an advantage?
Yeah, it can be useful for that. Or if they want to fly a specific line, for example, if you want to fly slower while climbing and you want to fly less fast while climbing, you also have the problem that the radio would beep incorrectly. And in that case too, it's obviously an advantage if you can just feel very precisely what's possible right now. If I don't have a day...
have, but also a lot with speed bar play and so on, then, if I understood you correctly, I'm constantly feeding myself misinformation with that. I back off the speed bar a bit, my glider climbs automatically because it's converting a bit of speed into altitude. The vario beeps and I think I'm turning into something, but there's no thermal there at all. So I just fly around it because there's nothing there. Are there techniques for how I should work with the vario better even without tech, or are there tips where you say, pay attention to specific flight techniques so that you can interpret the vario correctly and so you don't constantly induce misinformation yourselves? Phew,
difficult. I can't tell you. It always bothered me when I did a valley crossing and flew into a tube by accident. I actually couldn't center it. I tried and lost altitude every time. Because usually, a valley crossing is also associated with valley wind.
And if there's something in the final thermals, it's relatively narrow. And if you don't find it right away, it's just gone. I don't know how you're supposed to find it conventionally. That's why we built Tech. I can't tell you.
That means you go out of the speed bar and if it still beeps, you say, okay, here is the real thermal and not the one I've been faking, so to speak.
Yes, exactly. Right. You can really listen to the Vario, and the Vario simply shows you whether it's climbing or not. The real climb. The compensated climb. Not the climb because you're converting speed into altitude.
Have you become a better pilot because of your vario?
I think so, yeah. In what ways? It flies much more efficiently. You can center the thermals more efficiently.
For example, when it has very tight lines. On Sunday, it was very tight. It was also stubborn. And when you turn tight and then the Vario beeps and you know it's still going up, that helps. With a conventional Vario, because it's so stubborn, I think it's a bit more difficult. I do think it's easier with this one. That's also the feedback we get from the pilots. They fly longer, they do more flights and so on. That's the standard feedback we get.
If you look at your website, it emphasizes that, okay, the XC Tracer is a Swiss quality product. Do you really manufacture everything in Switzerland?
Practically everything, yes. Except the batteries are from Asia. You can't buy those anywhere else. And the black box with the zipper that the thing is in, that's also from Asia. And the rest is either from Switzerland or from neighboring countries.
But wouldn't it actually be much cheaper for you to outsource something like that? Like, just saying, "Come on, I'll find a Chinese guy who can assemble this for me and deliver the whole thing for a quarter of the price I'd pay to produce it in Switzerland?"
Yes, you could do that in principle. But we deliberately decided against it. We tried it once and found, no, that just can't be it. And we deliberately decided against it. We then built five small robots, for example, for the production of Max and Mini, which make production easier. And one of the robots is now in use at a supplier, also in Switzerland. And so it's actually manageable. Did you develop the robots yourself too? Yes, I built a small robot.
How can I imagine that robot arms,
...that arrange components on a circuit board and then solder them themselves? Or what does the robot do?
No, no.
There's glue, for example, for gluing the glasses together. Or we stick in beepers. Or we calibrate them. Such relatively simple tasks. But then you have a repeatable quality and then you can just do that in Switzerland. Sure, you still have higher costs than most competitors, but it's just how it is. Like
a lot of people
working with you? That's not many. It's just... me, Lisa, and sometimes Lisa's daughter helps out. And then we have external suppliers. So about three or four. We're not big. Small but mighty is our motto.
Small but nice. I mean, especially in the electronics field with such mini-quantities. You probably also have some special components. For example, the glider you use for the XC-Tracer and stuff like that.
Don't you also run the risk of quickly losing a supplier when you have such small quantities? Because they say, you're at a total of X, I don't know how many that is, but with the few units you have, you're actually not interesting to me at all. I'm taking you off.
I think it's the other problem. If you're not that big, you can always find some components. If you're relatively large, you immediately have a problem if you want, say, 10,000 pieces, because they just aren't available. You simply have to have good relationships with the suppliers. You have to have a good supplier.
We also had problems before not getting components, and we consciously decided against buying from a broker. Because when you buy from a broker, you never know what you're getting. You might get great quality, or you might get parts where only 50% of them work. So we said, no, we're opting against a broker.
Because if we buy bad parts, we're just constantly repairing things. We have unhappy customers. That ruins our business. It's completely pointless. And even if you build different components, you then have different microcontroller variants and then you have different software that you have to describe and so on. At some point, the whole thing just blows up in your face. We just said, no, let's not do that. Then we just wait until we have the parts. Recently, we waited eight or nine months until we could sell something again. But that's just how it was.
What did you do during that time, when you say we're buying? Did you just keep developing the software? Yes,
Development takes up most of the time. Production isn't really the problem. Development takes up most of the time.
So, you're constantly at your desk programming and looking at how I can make my sensor data fusion even better?
Yeah, not just at the desk. We also do flight tests to see what we want to do for future products. Building new hardware, trying out new variants. Testing GPS, for example. That's very complex. For instance, when we made the first Mini-GPS, we built over five—I don't remember if it was 20 or 25 prototypes—until one was good enough to be used. And Kriegerl flew with it for x half-flights. That wasn't actually planned. I just asked if he would test it for us because we don't get around to it, as we often don't have enough time to fly. And solar cells—whether the energy from the sun is sufficient—you only know that if you fly with them for several days or weeks. We simply don't, because we have too much to do. We gave it to Kriegerl, and then it went on for two or three weeks, and then he called to ask if he could use it for x half-flights.
I said, yeah, you can do it, but just don't publish any photos because there wasn't a product yet. It was just a prototype. We build a lot of tests. There are so many prototypes, we do so many tests. We also did C and FCC certification for the Max, Max 2, and Mini 5. Best in the radio lab. You pay 450 francs an hour to test those things. That's extremely expensive. Whether it's worth it is hard to say, but we simply tested the things to see if we could deliver the best possible quality.
One peculiarity with the XC-Tracer, from the small solar vario to the Max 2 with its small screen and so on. You never went for, let's say, building a large vario where you have route guidance, an FAI triangle assistant, and all that. Why not, actually?
Good question. We have to be able to handle it. That's one point. Another point is that the market—well, those are the people you hear about, the ones flying the long distances and so on. But I think the number of local mountain flyers is much, much, much larger. And if you pack in a lot of features, you can also invite a lot of trouble. As soon as you have too many features in there, you have too much trouble and then even more requests come in. Now you could add this, and that, and that. So we said, we'll do it the other way around. We try, true to the old Smart slogan "Reduce to the Max," to take out as much as possible and maximize the user benefit. Because then, if people ask you when you meet them at the trade fair, "Can I adjust this and that during the flight?"
Then I say, yeah, you can, but are you letting off the brake? No. Yeah, then why do you want to adjust something? Good question. And then the topic is usually settled. I think,
that's a very small market for the complex flight instruments. And by now, it's like, if you look at the X-Half, where almost everyone was using the Max 2 this year, they almost always fly with it. And if you want to navigate well, you basically need a second screen. And then we would have to offer a giant flight instrument, and I think the market is too small, it won't pay off. Too much programming effort.
for what you would end up getting out of it.
Yeah, exactly. That's the best part, when we look at making sure we have the best, most reliable, robust software possible. And I mean, success speaks for itself. There's a reason why almost everyone was using the MX-Halbs in 2023.
It's not because we built the lightest instruments. There are lighter ones out there. But apparently, the overall package was so good that most people chose it. Yeah, probably also
just because of the battery life, which I believe is among the longest of any devices I know on the market.
Yes, and above all, it's like this with us—you might see it on the homepage—we're set up a bit differently. We're not "blah blah blah, here's everything we can do," but we're rather conservative in our approach. And when we write that it has 70 hours of battery life, it's not 35 hours, it's really 70 hours. Yes, the battery life was certainly one point, but I also think the precision, the reliability, the good readability, the fact that it's compact and not too heavy. There are many different reasons why they fly around with it. Pilots don't just fly around with the XC-Tracer during MX-half-days, but also otherwise, when they fly competitions or when they fly privately, you almost always see the XC-Tracer in front of the lounge.
Do you actually live off the XC-Tracer today? Is that your company and your source of income? Yes, exactly. But if you Google a bit, you can see you pursued a few other projects in the past. One example, in terms of what you find, is Smartfish or rather a specific mechanical watch. Tell me, what's the story behind that?
Yeah, I'm just curious, innovation—that's my bread and butter, I'm good at that. Smartfish is basically, how should I put it, a new
I can't explain that.
In principle, it's based on the conventional laws of aerodynamics, just put together differently.
So it's called a fish, but it's actually an airplane?
Yeah, exactly, it's an aircraft. You can basically say there are bird-like fish, like manta rays and such, which move relatively slowly in the water. Then there are the eel-like ones, like the eel, which also aren't that fast. And then there are the streamlined ones, like the tuna, for example. And a tuna, when it's hunting another fish, also has to be able to swim in a curve to, yeah, the chase—you might have seen it already—its fish has to swim in a curve, and the one in the back can't waste too much energy doing that, so that when it finally catches one, it doesn't lose more energy from the hunt than it gains from eating. And the one in the front shouldn't really get tired, because otherwise, it'll get eaten too. And when you swim a curve like that, you also create induced drag, basically like with a plane—if you're flying a loop or a tight turn, the laws are very similar.
And then I tried applying that to aircraft, and it effectively looks very similar to a fish, and that has many advantages, but it's aviation—it's expensive, and it's very difficult to organize the money for it. We were unlucky because when it came out, it was just a month after 9/11 with the attacks in New York, and then it was practically impossible to organize any money at all for a project like that.
What advantages would a Smartfish, or any aircraft built this way, have over a normal plane? So, could this be used for passenger jets, or is it actually only intended for small drones?
No, it's basically intended for drones; for small two-seater training aircraft, it would be ideal. Because you can, well, I might have to say, what's the better car? A Volkswagen, a Bulli, or a Porsche? A Porsche obviously looks sleeker and goes faster, but if you look at the transport capacity, if you look at how much cargo you can transport from A to B with the VW bus, then it's much more efficient than a Porsche. And Smartfish is basically like a VW bus, but looks like a Porsche. That's also a bit of the problem,
you can transport goods or cargo from A to B. And that is much simpler than a conventional aircraft; if you pay attention to what sticks out from the wing of a commercial plane when it lands, it hardly has any more wing, but just some sensors and actuators and flaps and cylinders and, and, and. And you can leave all that out with the Smartfish. A unique aerodynamic concept? Yes, it's simply applied aerodynamically differently. So, for example, with a commercial plane, if you can leave out all those flaps, you can also save a lot of weight, that's one thing. The second is, you can save a lot of money, because manufacturing old things is expensive. The third is, you can make mistakes during design, you can make mistakes during manufacturing, you can make mistakes during assembly or during maintenance.
That's the third thing. And the fourth is, of course, there are gaps and irregularities all over the surface, and that creates additional drag. And now you can say, okay, maybe I can make the surface a bit larger, but leave out all those gaps again, and if I get rid of all those irregularities, I won't have any more drag despite the larger surface area. But if you have more surface area, under certain circumstances, you can carry more volume or more cargo or more fuel.
It's just thought of differently, but it's not—it's based on the established laws of aerodynamics and physics. It could be like that
will a Smartfish ever be realized? Or has the project just been shelved?
Yeah, we're still on it, let's see if there's anything left, I have no idea. I can't tell you, it's very difficult. I wasn't on the wrong continent, you'd have to go to the USA for that.
Your name first came up for me in connection with the paraglider scene when I first read about a project by Jin—the paraglider manufacturer Jin—who developed new gliders where we now also used new profile shapes or special profiles, so-called Equalized Pressure Profiles, EPP, which look a bit like Sharknose and such. And it said there in connection that, yeah, we developed that together with the Swiss company Schafroth, that's when your name first caught my attention. How did you get in touch with Jin and what did you actually do for Jin?
As for how the contact came about, I have no idea, it was a very long time ago, I've known him for a long time now, I don't remember exactly, yeah, he just asked me if I could do the air intakes for these profiles and then I did it, simply, he told me that the gliders have about four properties, sent me the profile, and then I just adjusted the air intake so that it stalls as late as possible, has as much pressure as possible, and still has a lot of pressure at small angles of attack. That's how it came about.
What do you mean by that, did you do the leading edge or the inlet opening accordingly? Is all of that simulation work you did, or? Are you still experimenting, tinkering with different models? No, no, no, that is
all on the computer.
And why can you do something like that? And Jin, who has been involved with paragliding for 30 years and probably has his own simulation software and everything else, why can't he do it?
That's a good question. Sometimes, when you're doing something yourself, you're too close to it and you don't see certain things. I can't tell you, but of course, I've done a lot of multi-objective diagrams and design optimization in connection with the Smartfish project, and I've learned a lot. I can apply that directly now, which he partially didn't do. That's no longer practically oriented; it's no longer about flying. You'd have to ask Jin. I don't know.
Are you still doing anything for some paraglider manufacturers or for Jin as well?
Jin gets in touch from time to time and then I do something, then I have half a year of peace, then I don't hear anything, then I hear something again. That's roughly how it is. It's more of a hobby, it's not a job.
That means he always comes to me, and then you say, I'll fire up my computer again and run a new simulation, and then I can tell him what could maybe be done a bit better.
Exactly, something like that.
Would you actually like to build paragliders today, like returning to your beginnings in a way? Would that still be appealing to you?
No, I don't think so. It was cool back then, it was fun, but I don't think I'd want to put myself through that stress again.
About, I don't know, ten years ago, we were invited to Korea by Jin and then it was almost like the old days again,
The knots, the flying, and then test pilots saying, "Here and there, you need to pull a bit harder." Okay, give it here, then the lines go slack and so on. I thought, no, that time is over. That's not really my thing anymore. It was super cool there in Korea, but yeah,
No. Couldn't that be done somehow differently? You, as, let's say, Daniel Düsentrieb, as a bit of an inventor, don't you have ideas on how paragliders could actually be designed better nowadays?
I think the many designers are very good and they have very good tools for how they develop the gliders. It is still the case that you have to fly a lot, because you can't do everything on the computer. You really have to go flying to see how the glider behaves in the thermals, if it turns well, if it stabilizes, if it pulls into the thermals, whatever. Does the glider give you a sense of security? You probably know that. There are gliders where you think, that just can't be right. And then there are other gliders where you think, I guess I'm okay with this? What's the difference? That's the criterion. How can you put that into numbers? That is extremely difficult. I think, around conventional testing, there's no getting around it. You can probably do performance calculations relatively well, but when it comes to characteristics, it gets pretty difficult.
That's my personal opinion.
Well, you've been watching the scene you're part of for 30 years. On one hand, the paraglider manufacturing scene, let's say, but also the general paragliding scene. From your perspective, has everything actually developed in a positive direction across the board?
In the past, it was more of a community sport. You'd go up together, fly down together, and meet up again at the landing field. Today, it's like you fly somewhere and then unfortunately just drive back. It's not the same community sport it used to be. That's one thing. The other is that, in my opinion, maybe too much importance is placed on performance—X-Contests, flying as much as possible, flying as far as possible, and so on. That's perhaps the second aspect that's a bit negative for me. The other thing, of course, is that the gliders have become extremely much better, they've become extremely much safer. You have possibilities that you didn't have a few years or ten years ago. Yes, it always depends on what you do with them. If you...
you have options, you can push them to the absolute limit, which might not always be a good thing. Or you can
use reasonably and then it's fun. So, purely in terms of flight performance, it has already...
well-developed. I don't personally think you have to fly around with your Submariner, but if you're on XContest, then you probably have to, because otherwise you'll get a few kilometers less distance than your buddies. Yeah, that's roughly it.
If you could change something in the scene in general, or initiate some kind of development, what would it be?
I'd say a bit less focus on performance. It's worth thinking a bit more about whether it's worth flying a few more kilometers, like maybe flying into a valley that's not so well-known. Or maybe the thermals are a bit too strong for my taste. I actually shouldn't be flying at all, but the others are flying too. Just say, maybe take a step back and enjoy it a bit more, a bit less performance, a bit more looking at the mountains, looking at the landscapes, having fun with it, and less counting the kilometers. I think that would be a good plan. How could you achieve that?
So how can we give the scene the impulse to say, "Come on, let's not push performance so high and increase the experience value elsewhere instead?"
I don't know. I'm an engineer, not a psychologist. I can't tell you that. It's difficult. I don't know.
How did it happen that you don't place so much importance on the idea of performance?
I was once at a Swiss Championship. I don't even know. I could do it once in the beginning at Salew in Geneva.
Yeah, if you just saw that they almost had collisions or even did have collisions in the air and stuff like that, it was crystal clear to me, that's not for me. I just thought it was way too dangerous. There were also a lot of accidents back then. At the PWC, there were serious accidents regularly. Yeah, and for me, it was crystal clear, that was the end of the matter. I wasn't really interested in it at all.
You just wanted to fly, but the numbers were irrelevant. Exactly, it was about having fun and that was it. If you're oriented towards having fun and pleasant flying and so on, have you actually ever seriously injured yourself while paragliding?
Yeah, I messed up about 4 or 5 years ago. I was working back then, testing the Mariometer. I wanted to land on the top. It wasn't actually a problem. But because I wanted to land on the top, I didn't look closely beforehand and suddenly there were cows. And even better, the right kind of cows, with big horns and all that. Then it became pretty quickly clear, oh, that's a really bad idea, you're not allowed to land there. So I thought for about 2 or 3 seconds about what to do. And then I made the wrong decision and said, I'll go down there anyway, but on the side. And then there was a relatively strong lee. And then I was almost with my feet on the ground. Then the Machal lifted up and there was a stall because I braked too hard.
Then I fell on my back and pushed my back out again. That annoyed me extremely because I hadn't had anything for over 30 years before that. That's one accident I had, and another accident was the summer of last year, so 1922. There, I simply wasn't paying attention. I wanted to try out some stuff there too. Then it slammed me down almost 5-6 meters in height. Then I had a total destroyer, or almost a total destroyer, a few meters above the ground. That's nothing, just really 5-6 meters above the ground is three-quarters of the glider gone. But that didn't do anything at all. I was still able to roll it off, the glider opened up about halfway again. But it was already wrecked. There was just too much wind, too strong wind. So at the top, there was too much wind, too strong thermals. The wind was, I think, around 30 km/h.
And there are houses and trees up front. And then combined with the thermals, it was just too much. What lessons did you take away from that?
Look at the weather report much more closely. That's one thing. The second is, I had agreed with Lisa back then that she
drives to the landing field. And actually, we were already in the weather reports. Flying to the landing field isn't really a good idea. You should land at the front of the lake, but they don't like it when you land at the front of the lake. I already did a flight in the morning and the wind was too strong. It was clear, only land at the front by the lake. And even at takeoff, it was actually, I shouldn't have taken off at all, because there were tandem pilots there. Normally they fly in Interlaken in all kinds of weather. And they were a bit hesitant. Then one went out. I didn't watch properly. And if I had watched him properly, it would have been clear to me that it was just a mess. I made the first mistake.
Take a small flight, test things out quickly, see if the software works and all that. That was a mess. Nothing happened, but you have to learn from it. Even small flights are important, they can be dangerous. Do you think a mistake like that wouldn't happen to you again? Well, you can't say for sure that you won't mess up again. But I am trying to fly more consciously. Yes, how should I put it? I think flying is probably more dangerous than we all think. But you can try to minimize the risks. And then, do I really have to fly a circle now? Wouldn't it be better if I also flew on the slope, with the strong thermals and so on? And only then, when I'm really above, above the Kräthe, can I fly circles. And that might not be such a good idea. I'm thinking about things like that more and more.
And try to fly more consciously to minimize the risk. But you can still fly over the Alps, over glaciers. That works perfectly. Do you ever fly without a vario? Well, there's always something new to try.
What are you working on right now? What's new to try out? Can you give a little preview of what might be coming in the future?
Yes, we're working on making the obstacle warning significantly better. That's one thing. Then, of course, we're constantly working on improving the wind calculation. The thermals assistant will also become significantly better. And currently, we're working on translating software, obstacle databases, and airspace updates in the future. So, some exciting things are coming.
So it remains exciting with XCTracer, and you'll still have quite a bit to do to further develop all the functions. Coni, thank you for the interesting conversation, the insights into both your creative journey and the technical details of XCTracer. Perhaps some of those who use it will now look at that little instrument they have on the cockpit with even greater respect and say, oh, there might be even more brainpower behind it and
Swiss tinkering ideas and all that, making sure it all works like that.
Yes, there's quite a bit more to it, but unfortunately, there's a lot we can't say about exactly what we do. But the customers have to appreciate it. That's the main thing.
That was Podz-Glidz episode 117 with Coni Schafroth. As always, you can find some further links in the show notes for this episode. I'm Lucian Haas and the producer of Podz-Glidz and Looglights. The podcast and blog are passion projects for me. However, as a freelance journalist, I also make a living from these offerings. That said, I deliberately forgo compromising advertisements, annoying paywalls, or subscription models for funding. Instead, I rely on the fact that, just as nature gives me lift when I fly, you as an enthusiastic listener and reader provide me with financial support. As a supporter, you can give as much as you want, or rather, decide for yourself what such an offer is worth to you.
A common donation amount is 60 euros a year, or about 5 euros a month. But no matter what you pay, every amount contributes to the sum that allows Podz-Glidz and Looglights to continue operating and developing for the paraglider scene in the future. You can find all the necessary information on how your donation reaches me at looglights.blogspot.com. There is a menu item called Fördern. Looglights is written as Lu-Glidz. Until next time. Don't fall from the sky. Ciao.