Slightly off topic, but I just wanted to point out that this technology they're working on is largely within reach of you as a hobbyist today. What's impressive is the engineering resources and software control systems they're adding.
But for less than $300, you can build a small 6" quadcopter from base parts, purchase a small control radio, and a pair of inexpensive FPV goggles+transmitter+receiver. Yeah, less than $300. China's amazing at this.
If you get a permit for it in the US (or ignore the law), $50 more gets you radios with miles of range. Your primary limit will be batteries (solutions exist, but bust the budget a bit, fixed wing drones are a lot of more economical here).
But there is nothing magical or special about Switzerland other than a concentration of interest. If you wanted, you could start competing with these folks and stand up in front of a shady incubator looking for your economic kickstart really, really fast. You don't need to fabricate parts to build prototypes and most of this stuff is mass produced.
The technology has certainly reached a point where you can quickly put product prototypes together, the current regulations around the world finally allow for commercial use of drones, companies are beginning to invest more in the many drone verticals (some of which are still being defined, like delivery), and communities are being formed to host industry collaboration, I think this is a great time for the "Drone industry" regardless fo where you are. I work for the Linux Foundation, I'm the Program Manager for Dronecode (dronecode.org), we host some of the most successful open source drone projects and give companies a vendor-neutral governance where they can freely collaborate. At our core, the project that runs this community is our autopilot, PX4, the project was started 10 years ago in a lab at (surprise) ETH Zurich by (now Ph.D.) @LorenzMeier (Github), the autopilot is being used in some of the products you see every day, has an ecosystem of tools, all of which is actively being maintained. You can find more info on the autopilot on px4.io
To @KirinDave point, yes this technology is available and you can start working on Drones now, the Drone community is very welcoming, I think its a side effect of having to apply technology from other communities, like computer vision.
> But there is nothing magical or special about Switzerland other than a concentration of interest.
You could say similar things about Silicon Valley. And yet most people agree that it’s a super important place for all tech stuff. So it seems to be a very relevant factor.
I wouldn't say so. SV has always had quite specific points that, for a lot of people, made it more appealing than the alternatives. There are a number of technology "interested" areas in the US including Dallas, Chicago, Boston, Seattle, New York and a lot more "lesser" ones that aren't SV.
Having lived in Silicon Valley, there is one magical thing: climate. It is warm year-round; Spring comes in January; You can go surf in the morning, go to work, and ski in the evening. Did I mention afternoons with 23 - 25 degrees Celsius in January?
The Wingtra in the article is a VTOL fixed wing, not as easy to do as a quad.
The other main drone from Switzerland is senseFly's eBee, again a fixed-wing mostly for photogrammetry.
The drone in a cage they talk about is I think flyability's.
All of those are drones for professional applications, not for hobbyists.
That market is pretty much cornered: wants nice videos/pictures -> DJI, a racing quad or just to pilot -> go custom/cheap Chinese.
In all of those cases, the value is either in software/product behind the drone where the hardware is just a platform, or something quite custom for a specific application.
As you said, you can't compete with the Chinese for generic hardware, as Intel, Parrot and others discovered.
Sure, but fixed wing drones are even easier to build, fly and (modulo the landing bits) land than quads. Even independent small shops are designing and shipping recreational performance fixed wing drones reusing drone parts (BlackSheep's Caipirinha 2 comes to mind: https://team-blacksheep.com/products/prod:tbs_caipi2_kit).
Quads just have the ability to hover, which is desirable in a lot of industrial applications. They're also a nice example here because they have the most expensive avionics software and hardware (clock speeds have hard minimums) and THEY are cheap.
For safety before flight, we at DronesBench http://www.dronesbench.it/en have devised and validated academically a single electro-mechanical parameter for the drone as a whole. If written on the plate at factory or just sticked somewhere on the frame, it would make testing extremely inexpensive through its entire operational life. Incidentally (shameless plug!), we are exhibiting at The Commercial UAV Expo in Las Vegas on Oct 1-3 with an eye to the STEM educational market worldwide.
A single parameter to benchmark seems like quite a big statement, given how varied the use cases are. Perhaps you can tell us more about it, and why it's a worthwhile measure?
Sure, that’s a quick presentation with the relevant academic references http://www.dronesbench.it/en/fomdrone/ . Our testing machines implementing that concept are not good enough for commercial purposes yet (the system is dynamic and some unavoidable interferences occur) but we are doing industrialisation tests in order to bring the statistical error on measurements below 3% and hopefully below 1% soon.
Correct me if i'm wrong, but I thought it was illegal to do this in the US: at least if you want to do it with commercial purposes. Make stuff illegal, and it won't happen. simple as that.
You do have to negotiate with the federal government for any serious business plan. This is true of pretty much all transit and shipping companies historically. In reality, the space isn't very well regulated. Not unlike Financial technology, there's a ton of room for startups to come and start disrupting traditional businesses. the overhead for entering into any of the pulley to find spaces where regulation is it play, is to make sure that you have lawyers who understand the issues on staff.
These things are fun to toy with. For greater transmit power (and a slew of frequencies to experiment with) I'd suggest getting a US general ham radio license, which are relatively easy to acquire. There's also an optional extra license but you don't accrue much in the way of added spectrum.
§ 97.215 Telecommand of model craft. An amateur station transmitting signals to control a model craft may be operated as follows: (a) The station identification procedure is not required for transmissions directed only to the model craft, provided that a label indicating the station call sign and the station licensee's name and address is affixed to the station transmitter. (b) The control signals are not considered codes or ciphers intended to obscure the meaning of the communication. (c) The transmitter power must not exceed 1 W.
They call video tech for model aircraft First Person View
Also FWIW here are the two easy study guides used for US Tech/General licensing that can be downloaded. You shouldn't have any difficulty getting 35/35 on each in ten minutes.
My experience has actually been the opposite. Most of the FPV pilots I've met are running their video TX at a power output that should require them to be an amature radio operator, whereas most of the guys controlling stuff on long range UHF actually know what they're doing and do so within spec.
Heh, yes, that's true about the FPV pilots. I'm pretty sure I've got an FPV camera kicking around here from China that actually transmits on the amateur bands, although it definitely wasn't marketed as a ham transmitter :). I also suspect that it'd be way out of spec, but I haven't actually hooked it up to the spectrum analyzer, nor do I use it. It's a curiosity that came in a kit a long time ago.
And I suspect the reason the people on long range UHF do so within spec is because, in my experience anyway, the ham guys get more and more grumpy as you go down in frequency. That video transmitter above operates in the 1.2GHz/23cm band, and no one (around here at least) seems to particularly care too much about it. Going downwards, the UHF (440MHz/70cm) band and the VHF (146MHz/2m) band are used pretty frequently for voice communication, so someone splatting all over the band would have people enraged. Going lower into the HF (<30MHz) bands, you've got people worldwide using those for communication, and they're more than happy to get the FCC or Industry Canada or whatever your local regulatory body is on your ass for breaking the rules.
You're correct. Sorry for being unclear. Also I genuinely don't know if you're even allowed to do video over HAM. I've heard stories about how local ham activists love to report folks for mucking up their precious bandwidth.
You definitely are! And my other two replies in this thread will probably shed a bit more light on both the video bit and the "local ham activists" bit :)
ATV is a cool thing! And, thanks to you, I discovered there's more to it than I remembered from when I last read about it.
The more common form of Amateur TV is SSTV[0] - Slow-scan TV. This runs in 3kHz (!!!) of bandwidth and typically gets transmitted on voice frequencies, since it can fit inside the max occupied bandwidth there. People transmit this on the HF bands, which can go world-wide! But... for FPV-purposes, it's pretty much useless. You're looking at something like 8-114 seconds/frame.
But I had no idea that there's people doing Amateur fast-scan TV[1]. That's pretty cool too! I have no idea how readily available light and portable equipment would be for the 70cm or 33cm bands, but there is a bunch of equipment available for the 23cm (1.2GHz) band. As mentioned in my sibling comment... it's not necessarily legal equipment, and it's kind of just... available... from random distributors in China, but it works OK and you're not all that likely to have someone complain about it in that band, unless you happen to be interfering with their attempts to do the same thing :)
Funny you mentioned SSTV. I'm putting some SSTV software together and have an SSTV contest award hanging on my wall. It's a lot of fun. Fast-scan (what I always think of when I hear ATV) seems pretty extreme to me though. :) Wiki seems to indicate it's used for drone work...or could be...
Huh! I suppose you could do SSTV decoding with a sound card eh? Hmmm, a bit of research suggests there's a 2m SSTV station in town here, although it's not clear if it's still alive. I might have to spend some time fiddling with GNURadio tonight and see what I can whip up.
I kind of want to pull out that cheap Chinese analog FPV camera too and see how disgusting the spectrum on it is. I have pretty low expectations (camera + tx + rx for $20? yeesh...)
Do you know off hand if the toy copter cos are looking to take advantage of the affordability of this kind of tech, or are thry just too old fashioned to care?
You can absolutely buy a tinywhoop or something like that. You can even wire them for FPV, which is a cool way to experience life in your house as a fly.
But the prices aren't just affecting the floor models. DJI's had to move into the $300 range (I literally just bought one) because of steep competition in small shops making custom drones and other vendors pushing the cost down as various manufacturing houses spin up.
The Cheapest one I saw was the Spark at 399 --looks great but looks like best experience would require doubling that for the goggles as well. Still; looks enticing!
If you want a mavic, just buy a mavic, building something with this level quality / reliability is not worth the effort.
The advantage to building your own is that you get a different type of quad -- a racing quad which feels more like flying a porsche (feels like a race car) than an mini-van (e.g. mavic, reliable but super fun to fly).
If you want to build a racer, rctimer has some of the best kits (high quality parts, well thought out design). Here's an old build guide from my blog, you would want to get a newer kit with an omnibus-based flight controller, but the idea is the same: https://nathan.vertile.com/blog/2016/04/24/rctimer-u210-mini...
I'd recommend going to the youtube Channels Rotor Riot and UAV Futures (and possibly Mr. Steele) if you want to get a nice series of well-sourced introductory materials by veteran professional drone operators.
I'm by far an amateur, and don't really have enough sim hours to fly the drone I'm just about done assembling.
But: the aerial camera gimbal is by far the hardest part to get right, and why vendors like DJI charge what they charge. Software exists to make it work, though.
Despite Switzerland's reputation, and even arguably better conditions, relatively similar countries like Israel and Sweden have been running laps around Switzerland in the last 20 years when it comes to startups (possibly excluding pharmaceuticals). It might be, but I don't see why it would be different this time.
Sure (debatable, actually), and what's your point? Many countries have been running laps around other countries in many industries, what does it have to do with clusters?
You are saying that it is special because of the clustering, so I am wondering what indicates that? If you look at Sweden or Israel it is fairly obvious that they are successful in at least half a dozen areas. You can trace the path from one area or company to the next.
In Sweden Markus Persson who made Minecraft used to work at King. Jacob de Geer of iZettle was the first employee at TradeDoubler, which was founded by Martin Lorentzon who would later be a founder of Spotify. Daniel Ek of Spotify used to work at another gaming company and of course the music industry is large in general. Zennström who started Skype used to work at Tele2, a telecom operator. Klarna was started with help from people from Bluetail, which used Erlang in turn made by Ericsson. Etc. Etc. (This is just off the top of my head and all these companies had large exits or were already established).
That is what a successful cluster is and what makes them hard to replicate. Without this dynamic it is just a bunch of companies next to each other. So why doesn't Switzerland seem to have that?
parent is saying that clustering is magical in the relevant field of conversation -- here, Switzerland:drones, not startups or companies. In fact, Switzerland is a demonstration of lacking the clustering for successful companies and startups, just drone enthusiasm independent of that, and showcasing clustering as kind of a magical yet incredibly powerful thing, proving parent's point.
As I said a cluster is powerful when it produces these effects. Otherwise it is just a bunch of companies with nice titles waiting to go work in banking. Countries like Sweden have consistantly shown that their clusters are powerful by producing these effects. Switzerland hasn't so why should we assume that this "drone cluster" is powerful and not just another empty claim?
> You seem to think that technology equals "venture capital that shows up in the tech news".
The companies I mentioned all have had great real success. They have hired thousands and thousands of people and ended up selling for billions. This "drone valley" is exactly the "venture capital in the news" that you are talking about.
> you don't seem to know much about Switzerland as well.
Why wouldn't I know about Switzerland? We have already established that Switzerland has a great reputation and good conditions. Sweden and Switzerland have a lot of industries in common. Like pharmaceuticals, construction, chemicals and automation. The Google office is well know as is banking, watches and other large brands. ABB is even Swiss-Swedish and of course Kamprad used to live there. And yes, I have been to Switzerland. At least a half a dozen time over two decades and more counting stopovers in Zürich mainly going back from Hong Kong.
This just doesn't have much to do with the topic at hand. Many places have great technology, but fail to break into new areas. Which is what I have been saying that Sweden is probably worse on papper yet far more successful in this regard.
> Or why Switzerland is one of the world leaders in research?
Because of good schools, high salaries and dedicated people. Which is perfect for people who want to have a good career. Not necessarily for starting successful companies in new industries. If you haven't noticed it is rarely "the best" people who start startups. They have too much to loose. It is the restless, the dropouts, the migrants etc. If you read my previous comments you can see that I am pretty bearish about Sweden's future for startups as well.
My question is simple and I have asked it repeatedly i.e. what indicates that this "drone valley" will go anywhere? Don't blame me if you can't come up with an answer.
They have to work out the air traffic issues. I work for an organization that does mountain flying in SAR scenarios (airforce, helicopters). We can be re-tasked mid-air to a SAR scenario at any time. We need to know within minutes, seconds, that the low level airspace is clear for us to operate. Can these drones avoid large helicopters on their own, or do we need some system to ground them all automatically?
This matters for us in particular because SAR can happen literally anywhere, not just in restricted airspace around cities or airports. We really do operate at less that 100 feet. We have landed in hospital parking lots. No notice on the ground. Just suddenly a huge helicopter is parking itself beside all the SUVs. That's what emergency means.
So a large number of metal drones is a real safety issue.
This is a very valid use case and something any UTM solution will have to solve. Currently the industry is hoping the FAA will create an uber solution and avoid “patchwork” solutions with state and local involvement, but that will never scale. Local administrators will need to be able to geofence areas in realtime. This will happen but it will take time to roll out and will require local, state, and federal involvement in a manner similar to auto transportation but with some unique aviation aspects.
>> require local, state, and federal involvement in a manner similar to auto transportation
Or, we can do it in a manner similar to aviation. State and local governments have no say. FAA rules govern aviation, with preemption making sure those rules are not complicated by local authorities. Our helicopters cover a huge area during normal operations. During cross-country transfers to things like airshows they operate across the entire continent. Local, even state, rules won't be welcome. It all has to come from the FAA.
Don’t believe an solution based purely on current aviation rules will work for the scale drones will operate at. Firmly convinced that the 200-400’ feet will more closely resemble class-A airspace than class-G and will involve local. Rules will be standardized by the FAA but the administration will have to involve local and they will have a say in restricted airspace, routes, etc. A VFR free-for-all with drones will be a mid-air nightmare and no magical UTM solution will solve it. Don’t disagree that the FAA has preemption, but the current aviation model won’t work in the future and the FAA knows this. They are pushing for more local involvement and it seems to primarily be industry pushing back. Additionally, it isn’t settled law where property rights end. Causby said 87 feet for sure but possibly higher. [1] Still a lot to be worked out how drones will integrate in the national airspace.
Local input on rulemaking is fine, but local towns dont get to promulgate rules as they do speed limits for cars. We cannot fly if every county gets a direct say as to who can fly where on sunday afternoons after labor day. They want to. They try. I field the calls, sometimes from local cops. Federal premption protects us.
Interesting to see this on front hn, today I’ve packed everything in London and left for Lausanne, arrived few minutes ago with family. Although innovation activity wasn’t the reason we’ve moved here, it’s very interesting to hear stories like this, there seems to be something good happening with research, gov backing etc. I’ve noticed some good research in areas like cryptography (zero knowledge proofs), formal proofs and now robotics/drones; I know government is also running couple of projects backed by blockchain, quite open minded, not seen much of this kind of interest from other European governments (except Estonia maybe).
Having spent some time on business applications of drones, especially imagery, I am actually quite bearish on startup application of drones.
Even if there are some locations that accept innovative experimentation with drones (e.g. Switzerland or Japan), the majority of countries (e.g. the USA with part 107) are quite restrictive. Because of severe restriction on autonomy, all large-scale applications are not legal.
Not many people are investing time and money into building new drone applications, as it's uncertain it will be legal in the near future. Without serious $ behind it, drone regulations will probably stay restrictive. If you are thinking about Amazon - I heard from multiple people that last-mile delivery is actually more feasible with ground robots.
This will start changing as technology improves. Specifically it will change as applications emerge where these things provide a concrete economic benefits that offset the risk. Basically when that happens, money starts flowing to regions where the legislation is more favorable and more innovation happens. At some point there's a tipping point and legislation adapts due to pressure of local parties missing out.
E.g. the agricultural sector in the US has enormous power on this front. They already use lots of drones as well as e.g. helicopters to dry cherry trees, crop dusting planes, etc. That business is quite expensive and an obvious target for replacement by drones. When technology gets to the point where legislation is the only real obstacle, there will be lots of pressure for legislation to change. I'd say that's a question of when not if.
What will need to happen on the safety front is for drones to play nice with regulated airspace and controllers/pilots in the area. Mostly drones currently don't have stuff like ADS-B or transponders on board. Given how cheap and light stuff like that is getting, I think it's a matter of time before there will be some drone specific versions of that that allow controllers to monitor any drone activity in their area. I imagine stuff like that won't be optional for any commercial operations. Mostly drones are a danger only because they stray in the path of other planes. If there's collision warnings, controller advisories, etc. things get a lot safer.
> E.g. the agricultural sector in the US has enormous power on this front. They already use lots of drones as well as e.g. helicopters to dry cherry trees, crop dusting planes, etc.
There is a bit of a chicken and egg problem here. As long as there is a restriction of a single drone per pilot, using airplanes and helicopters tends to be cheaper for most use cases.
Only under a closed world assumption. If e.g. Mexican and Canadian farmers start seeing real cost reductions by using drone technology there will be some very real market demand to sort out any legal obstacles for doing the same in the US. Basically this article is about the Swiss being a bit more pragmatic here. It's probably not the only part of the world where this is the case. I imagine the Chinese are very eager to get a head start in this market as well.
So, the only effect of this legislation will be that likely successful companies in this space may not necessarily end up being US based companies and that the US will end up being late to the game. That in itself might be enough incentive for this to get fixed sooner than later.
In the same way, self driving cars are being invested in despite legal obstacles that currently make operating them illegal in most places.
This is similar to the landscape for self-driving cars, but I don’t see how it’s an impediment. Maybe someone can tell me where I’m wrong:
Even if only one country in the world encourages these kinds of autonomy, isn’t that adequate for the development phase of the tech?
If, say, India permits autonomous drones broadly...
1) That market alone can sustain all the research and initial marketing to sustain the industry to where there are mature product offerings.
2) Once there are mature product offerings, then any country that allows them will put their citizens at a competitive advantage.
3) Since permission is more easily granted than revoked, that advantage will spread faster than it contracts.
4) At some point there is a tipping point where money is clearly left on the table in banning countries, and the bans become a clear disadvantage.
5) Autonomy is permitted in all but a few markets.
I honestly don’t even see how that process would even slow growth, let alone halt it.
People act like these technologies are ready-to-roll and regulation is an impediment, but it seems like regulators can move pretty slow and still be fairly close to the growth curve the development and marketing timeline is capable of.
I work at Picterra[1], a swiss company doing machine learning on drone & satellite imagery.
From discussion with a number of drone service providers, it seems to me the restrictiveness really depends on the context.
For things involving flying a drone over a crowded area, it's definitely hard to get authorization and we've seen projects being abandoned because of that. So yeah, drone delivery in cities might take some time before it is allowed.
On the other hand, most of the commercial applications I've seen are around using drones for mapping. So basically taking pictures and then processing them (create orthophotos, build a 3D model, detect objects, etc...). This usually doesn't involve flying over crowd and for a number of use cases (farms, construction site, security) the fly would be over a private property. A lot of people are able to do this without any legal restrictions.
From what I know of the history of the swiss Drone Valley, the two biggest success are sensefly [2] and Pix4D [3] and those are in the mapping business.
Two of the booming startups would be wingtra [4] and flyability [5]. Wingtra seems to be mapping-oriented as well and Flyability develop drones that can fly inside. So a lot of the swiss drone valley is really into drones that can take pictures for mapping & inspection and I think this segment will continue to see massive growth in the future.
> On the other hand, most of the commercial applications I've seen are around using drones for mapping. So basically taking pictures and then processing them (create orthophotos, build a 3D model, detect objects, etc...).
In my opinion, the majority of those applications are better served by either CCTV or manned aircraft.
Mapping use cases - If the area is big enough, manned aircraft mapping is much cheaper than drone mapping. Typical airplane ortho is around the 10cm/pixel, so it is sufficient for the majority of use cases. E.g. Check our nearmap. If the area is small, the machine learning is not going to make lots of $, as it's better to just use humans.
Monitoring use cases - e.g. "I want to detect if construction workers are wearing helmets" or "I want to detect poachers in the area". You can provide more business value by doing computer vision on CCTV camera.
Wingtra, the firm in the article with the helicopter/airplane-like drone really made some headlines in local newspapers and rocked some prestigious startup rankings.
Switzerland is clearly a startup hub for B2B companies and high/deep tech. I guess, with only 8 million citizens, we can't go B2C and don't have any chance to compete internationally other than "improving" B2B. Especially recently Swiss firms invested a ton into automation after the "Frankenshock", a day in February 2015 when the Swiss Franc appreciated by 15% to the Euro overnight and hence endangered exports on which we rely on.
(Content marketing: If you are a software engineer with a EU-memberstate citizenship, and want to move here, there are many interesting, well-paid coding jobs in Zurich. Ping me at iwan@coderfit.com, if you want to get more info.)
Also, for folks interested in general "drone" (aka research/commercialized/militarized robotics) news, the http://dronecenter.bard.edu/ newsletter provides fantastic weekly coverage.
Thanks, OK so this definitely more of a "logical" valley :)
Besides the mentioned drone development is there much other tech and start up activity in and around Lausanne? I feel like Zurich seems to get most of the attention for tech in Switzerland.
Which, considering the small size of our dear country, basically makes "the stretch of land between EPFL and ETH Zurich" more or less the whole of Switzerland ;)
Hmm... all small scale projects. Point to point Drone delivery as killer application is still not possible in Switzerland. Need special permission for each flight/route.
If you're thinking of an Amazon type delivery (and not, let's say delivery of industrial goods, medicine, ...), the audience would not be there anyway.
Notably, major online retailers trialed same-day delivery (usually via bike messenger) within major urban regions (e.g. Zurich) for a moderate charge. Customers generally agreed that it was a great option, but no one used it, so it got canned.
Drones are not only quad-copters, they can be fixed wing too and then the noise goes way down. Also much more energy efficient to fly with fixed wing than eight propellers. And higher speed too.
But for less than $300, you can build a small 6" quadcopter from base parts, purchase a small control radio, and a pair of inexpensive FPV goggles+transmitter+receiver. Yeah, less than $300. China's amazing at this.
If you get a permit for it in the US (or ignore the law), $50 more gets you radios with miles of range. Your primary limit will be batteries (solutions exist, but bust the budget a bit, fixed wing drones are a lot of more economical here).
But there is nothing magical or special about Switzerland other than a concentration of interest. If you wanted, you could start competing with these folks and stand up in front of a shady incubator looking for your economic kickstart really, really fast. You don't need to fabricate parts to build prototypes and most of this stuff is mass produced.