Another reason (the article didn't delve deeper into) is that the drone hardware market itself consolidated and non-Chinese manufacturers just got outcompeted and wiped out.
Every day on cnn, cheesy computer generated images of drones delivering produce to urban food desserts. Drones! Package delivery is just around the corner! Look, a drone taxi service! They have one precarious prototype and no basis in economic reality. But their leader wears a turtle neck and drinks Raw Water and only he is able to understand how the economy of tomorrow works. Silly us and our pragmatism. Turtle neck man sells and in the end he gets away with millions and the rest of us stand in a crater where a billion dollars of intelligently invested money could have made a real difference in people’s lives. Where will turtle neck man show up next?
Subtantiating that CNN are pretty bad on this front.
It seems to me that if anyone should be pouring scorn on unsubstantiated techno-hype it's all of us here. People doing startups have the most to lose when snake-oil salesmen have their way and dominate the discussion. When the market then implodes we're left without funding and support for real technology that makes the word better. We get the reputational hit for BS we were no part of. Forgive us if those of us who are serious about technology and reality feel just a little emotional about that.
It needs to be called out by the knowledgable every time.
Especially when everyone who posts regularly here knows it to be true. Hype for snake-oil-tech is poison for us as a community, if indeed that is what we are.
People obviously disagree with the above. I must say I am utterly baffled as to why. Snake oil hype is poison for your startup unless you're in snake oil yourself.
"For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled." --Richard Feynman
You've probably seen that before too, or at least should have. Still worth repeating, often, in hype bubles of which there is clearly one surrounding drone delivery. At least I'm not trying to do something genuine, honest and real in that space. Must make things borderline impossible if you are.
Anecdotally, I never got into drones because of the batteries. Everyone I met that had them absolutely loved them, but had like 4 batteries they needed to keep charged for about 15 minutes of flight time.
And quite a few tried pushing that a little too long an ended up with crashed drones. So my half cent is that batteries need to get better before people like me even consider them. Mentioned 0 times in that article.
Alternatively, drones need to get efficient enough to be solar powered, which is maybe the intent behind the acquisitions by PrecisionHawk Inc.
This. I recently worked with a company that produces drones among other things, and spoke with the head of research there.
More or less, they said they had given up on putting money into drone R&D until battery technology caught up, as they had plenty of ideas that they believed in but were unfeasible due to battery life.
As RC vehicles {car|helicopter|plane} get larger they reach a certain mass where liquid fuel - usually a petroleum - is preferable to battery. Why is this not an alternative for drones?
Given the line of sight limitation for drones, gas power does not add much value. Batteries are usually sufficient to map/photograph the area within range of the operator. Given this legal limitation, it is often more economical to hire a small plane like a Cessna with an avionics package and a 400 mi range.
It's hardly $30k to buy a gas drone. You can buy a Cox .049 engine for $50 or so.
You're looking at a fixed wing aircraft though with one of those. What's made the quadcopter drone possible is the high torque and instantaneous throttle response of small electric motors, combined with fast control electronics. The conventional way to fly a small helicopter is just like a full sized helicopter, with a main rotor and a mechanical swash plate.
With the right control electronics, these can easily be made into drones, but people tend to not like them as much anymore because they're mechanically complex and finicky. The electric motors are simple, and when one goes bad you just drop in a new one.
Could you couple one of those engines to a generator, and use that to power the electric motors? You'd need a battery or supercapacitor to handle change in throttle demand that outpaces generator response too.
Weight is the limiting factor. Try to find a generator that can output up to 400amps at 25V (what my big drone draws) and yet weighs less than five pounds including fuel.
I think that’s a more impossible goal than increasing power density in batteries is.
I've watched people fly RC airplane and helicopters with gas engines. They still exist, and there are air fields for them. I've seen them since the 90s as a kid. They aren't expensive but, are bigger and require an actual landing and takeoff area.
Controls are more complicated if you're not using electric motors where a microcontroller can easily adjust the RPM. Someones already linked a hybrid drone with a gas powered generator but it's not cheap. In high volume it could probably come down a lot but I don't know how much demand there is for it.
The electric motors used can change the speed of 4+ independent propellers extremely quickly.
This is hard to do with petrol engines, without costing a lot of extra weight or mechanical complexity, in a system where any single failure means a crash
I wonder if there's some simple variable transmission that could be electronically controlled to deliver power from the motor to the propellers in highly controlled fashion.
A generator and four motors. Efficiency would probably be around 70%. It's about 90% in diesel-electric locomotives where more weight and money is spent on efficient conversion hardware.
This might enable electronic control witout using any motors just one simple coil per propeller. Energy generated through this way of control could be stored to power onboard devices.
I have a DJI Spark. It's loud when it's close but as soon as it's a few hundred feet away you can barely hear it. I primarily use it in nature and usually use it at 200-400ft elevation. At 200 ft it's quieter than standing next to a running car. At 400 ft you really have to try to hear it.
It's loud when it's close but as soon as it's a few hundred feet away you can barely hear it.
If you're in an open area. And if there is only one drone.
A couple of years ago I was in a serene canyon in Utah enjoying the peace and watching nature when three drones decided to scare away the bighorn sheep.
People don't hate drones because they hate drones. That's why many seem to like the idea of Amazon/UPS drone deliveries. They hate drones because of what the drone pilots do with them.
"Quieter than standing next to a running car" isn't a good criterion for subjecting strangers to noise in a public place they're probably visiting at least in part to get some quiet and seclusion.
I don't mean to be a dick; I've been flying R/C for ~25 years and I know how difficult it can be to find fun places to fly and how frustrating when other people get in the way of that. But I also really appreciate quietness and because I live in a city it's a rare gift for me when I do get out into nature. Please have some consideration for others and don't unilaterally decide what noise level is acceptable for them unless there are already significant sources of noise present or you're really sure nobody else is around.
Yeah but mostly on BLM land, and that's kind of what it's for? As long as the local USFS rangers are writing tickets, they mostly stay out of national forest. I'm never seen them at national parks, wilderness areas, etc.
I was trying to think of something to compare it to that others would relate to. A car wasn't a good example since the dB level varies significantly. I was trying to make the point that it's quiet at that distance. If you were having a conversation with someone you wouldn't notice it.
Maybe you have to try to hear it, but I spend a lot of time in large public parks and, without fail, hear drones flying overhead.
This just started in the last year or so, where virtually every time there's a drone off in the distance.
I don't have great hearing either (especially in one ear), after playing some musical instruments that you hold close to one ear to play. Those things are just loud and stand out when the other noises are the wind in the trees, birds, and the rustle of small animals in the underbrush.
I imagine if you get an FPV headset it's quite a fun activity.
If people can spend thousands of hours playing flight simulators on the computer, surely there's a lot of replay value of doing the same thing with an RC quadcopter through a forest or around your home/neighborhood w/FPV.
I made my own flying wing for about $300 all in (the controller and goggles are usable with any other open-hardware drone, though). It's really fun, I recommend it, and it's not as loud since it only has one motor.
No, it is a fairly stock one. However, tell you what, if you want you can follow me on https://www.instagram.com/stavrosware/ and I'll document it there (I'm rebuilding it).
Wow, that is impressive. I thought I would have to spend around double or triple the amount to get started with this. Would be really interested if you could share some details on this.
Certainly. I am also rebuilding the wing and another commenter asked for info, so I can post photos and details on https://www.instagram.com/stavrosware/ if you want.
I bought a Taranis QX7 remote controller (~$100), which is amazing and I love it and you can use it with any other RC vehicle you want. I even hook up the receiver to my hardware projects so I can control them remotely.
You also need an FrSky XSR or R-XSR receiver (~$10), a Beitian BN-220 GPS (~$10), a motor and ESC (be careful, get a plane ESC, not a quad ESC, the plane one is larger and has better heat dissipation, ~$30). You need two small servos for the elevons, those cost about a dollar each. You also need a good soldering iron (the TS100 is $35 and is amazing.
For the controller, the Kakute F7 is fantastic and costs $45ish. It can fly in modes ranging from manual to completely autonomous (you get a Google Maps interface and mark where you want it to go and it just flies to those points). This may be a bit overkill, as it's the best board around, you can get a cheaper one (or maybe one that is made specifically for planes) no problem, but it's only $10 cheaper.
The FPV goggles can cost around $30 for some cheap ones, adjust those to taste.
I think that's it for components, although I should probably do a proper writeup if people are interested. There's much info out there already, but it might not be concentrated in one place. Let me know if you have any more questions!
Anecdotal, but the FPV drone pilots I've seen are much more responsible and aware than photography drone pilots (I'm a part of both communities).
Photography drone pilots will do all kinds of stupid shit that horrifies me, like fly near people, near obstacles, under trees, in crowded places, etc etc.
I blame it on the fact that a Mavic is super easy to get and requires no knowledge, whereas it took me two hours of reading just to figure out how to start my racing drone.
I usually get into stuff like this for professional reasons. However, as someone in the aerospace industry, it’s not very common that we go through the hoops to get a drone out at the airport. We’ve talked about using one strictly inside of our hangars, but at that point a gyro does just as well.
Have you ever been to go karts? To me flying a drone is much like that. I have never had an expensive drone only flew the 50$ or less drones I have gotten my young children. Flying around the yard avoiding things trying to land in different spots is a heck of a lot of fun with no real purpose. I can only imagine the countless hours of fun I would have with first person view flying. So for me I like drones just for fun. Though I do not fly every single day, every once in a while I will go out for a fly. The drones with the ability to take pictures is a lot of fun. A drone getting away on you and almost crashing but adrenaline kicks in and you run after it and save it in the nick of time is a thrilling experience. This some of the best fun I’ve had playing with my kids toys.
I've had a couple of <$50 drones which I've played with a bit. It was fine for the money--though very fiddly, at least the ones I've had. Hard to actually use the camera.
Fun enough for a few hours punctuated by recharging but not really something I see myself spending $500-$1000 on. I don't have kids. I can see them being fun toys. I spent many hours as a child playing with balsa wood gliders for example.
I feel much the same way. The one I have came with 2 batteries and with a couple more I think you could continually swap out and it only takes second with no tools required. The charge time was under half hour. But like you say I don’t know if I could push myself to pay 1000$ for a drone. I mean with one like that you can do a lot more and the range is a lot farther but at that price I would be terrified to lose it if it ever left my sight. The 50$ drone I push for my own adrenaline’s sake, over the water, through the trees, over the death spot if it dies. But if 1000$ falls into the water I will be swimming and I don’t like getting wet!
Longer flight times are possible, if an engine is used instead (plus servos to change prop pitch). I'm just not sure why it hasn't been pursued much (can't be the noise - electrics aren't exactly silent)...
People have used fuel-powered drones extensively before Li-Po batteries got good (including me). My take on why they haven't taken off as much as the modern quadcopters: they are quite a bit scarier.
The main two types of flier before quadcopters were fixed-wing model airplanes, and helicopters. They are both riskier to use than quadcopters: fixed wings can't be stationary and need a large area to fly around, while helicopters have large dangerous blades. This is why either of these are less attractive to the mass market.
On the flip-side, I think converting a quadcopter to use an engine is expensive because controlling 4 blades at once is not as simple as with 4 electric motors. Either they need to have 4 gas engines (expensive, not as responsive as electrics) or they need to have a variable pitch mechanism (now you need an intricate/expensive mechanical system of levers and servos to transfer the control).
It has, just in the form of Helicopters. Its been known for a while that moving a lot of air slowly with a bigger blade is more efficient than moving less air faster with 4 props.
The issue is that stability and control is easier to do with 4 blades rather than with one that has a swash plate.
It depends on how you use them. I travel full time so I primarily use mine for nature photography. 15 min is usually plenty of time and if not I have a 2nd battery that takes no time to swap. It would be nice to have longer battery life but it's not a deal breaker.
The only cases where needing more time would be a big deal would be for filming something that lasts longer than the battery.
With petrol’s energy density, why not make gas powered drones? Gas engine → generator → small battery (backup power) → electric motors. This is what manned drones (still in development) do.
Heavier fuels (kerosen, a/k/a jet fuel, diesel) have equivalent energy density to petrol, but are far less flamable. You might get a small fire, but not an explosion. And there are a lot of RC aircraft already.
They are noisy, however. Drone buzz now is simply from spinning fans. Adding a reciprocating motor to the mix is all the louder.
The alternative is lighter airframes and solar power. For quads / octos, probably not viable, but for a prop-driven fixed-wing craft, a very small amount of thrust to counter gravity could be quite effective. Expanded polystyrene or similarly light body, solar cells over all horizontal surfaces, and an ultra-lightweighted controls and imaging payload. That would allow for long-duration flights at least in daylight.
The low-energy use would also make battery-powered night flights more viable, certainly beyond 15 minutes.
The motor to run the drones was expensive till they became mass produced and dropped in price. Same for the batteries. DJI Drones aren't exactly cheap either.
also I believe if you want to be pedantic, if it runs on combustion it's a engine. if electric then a motor.
Given the line of sight limitation for drones, gas power does not add much value. Batteries are usually sufficient to map/photograph the area within range of the operator. Given this limitation, it is often more economical to hire a small plane like a Cessna with an avionics package and a 400 mi range.
My brother used to have RC airplanes. Gas is messy, and starting even a single prop was a massive pain. Motors are also less laggy, and probably easier to precisely control than engine throttles.
Now, that's for small scale drones flying 15 minutes. For, say, commercial airplanes that can fly 18 hours, energy density becomes a major factor.
For me, it’s the FAA notification requirements that seem daunting. It looks like the situation may be better than it used to be, but it’s still confusing at a glance.
I never got into them for the same reason I never got into football or (before I had to for professional reasons) Linux - everyone who is into it turns it into an evangelism project.
I’m sure I would love them if I tried them, but I don’t want to be seen as the type of guy who owns a drone (like how I don’t want to be known as the guy who uses Linux or the village Libertarian, even though those words might describe me reasonably well).
We are now entering the trough of disillusionment. In a few years some real novel and amazing companies will emerge and drones will find their useful place in society.
This. Now is a great time to innovate drones. The developments in FPV drone tech are happening fast and it's mind blowing. For example, 50sec into this clip: https://youtu.be/-AL5S3g03qw Stay tuned, drones are just getting started.
No, we have had great pilots and editors for years. The new gen FPV hardware has low latency HD for the first time this year, with better dual gyro flight controllers on faster chips, faster ESC chips and new FC & ESC software for superior filtering. Lots of developments like Neuroflight firmware are also very new, it's all amazing when you dig into it. Also new this year are tiny fpv racing cinewhoops that are 4k capable and fit in the palm of your hand, and fly around 100mph...
https://github.com/wil3/neuroflight
Most of the good ideas are either massively competitive (aerial inspection/ag support/photography), a pain in the ass due to safety regulatory issues (drone delivery, passenger flight), or just bad ideas overall (consumer camera that flies around and tries to follow you).
There are occasionally bursts of genius in the field though. A few years ago I met a guy (Sergei at Fotokite, I believe), who was developing a tethered drone for photojournalism. Solves many problems at once: you don't run out of battery, you could control the drone via a stick or just pulling it around by the tether, people don't hate the drones as much when they can see the operator. He demonstrated it by deploying it in the reception area of an office building, probably 15x15ft. Now apparently they're developing a tethered drone for first responders that automatically deploys and gives you a 150' vantage point for situational awareness.
BTW: No relation to Sergei or Fotokite, just impressed with their tech.
10Mbps DSL means you're two miles from a central office. For $20,000 per mile (maybe less?), you can get fiber run to your house. A friend actually did this.
Ubiquiti sells you a pair of AirFibre gigabit microwave links for $3000. I'm sure you can rent a little tower space with fiber internet for a very long time for $36k.
This is the way i will go if i do anything. Got a quote of $9k for an 80’ freestanding tower to get over the tree canopy where i live. Then it’s just hardware costs plus probably $500/mo for gig internet plus antenna placement in town. I could easily sell 100M to 4-5 neighbors for $100/mo.
I got it quoted last year, $32k for fiber to the node (no CO in our case) plus loop to nearest exchange. Their network is terrible and it didn’t make sense to rely on them for IP.
I once had 4Mb one mile from the CO. Then water got in the line and I was moved to a pair that only served 2Mb. That's the beauty of Frontier's customer service.
Definitely an issue in our case. This is centurylink and their subscriber IP network is dog dookie (qwest handles the core out here and seems ok). My $32K quote included a loop to a public exchange in the local metro area where I could pick up internet from a huge list of providers.
I’m fairly familiar with one of the businesses in the article and it’s just fascinating to me. They took a solid product that would’ve absolutely made a perfectly good, sustainable business line, added a ton of VC money, and got utterly swallowed up by buzzwords.
My only experience of VC money was during the dot-com bust. Worked at a pretty successful start-up that was good at what it did and was growing organically slowly but nicely. VC money came in and suddenly the CEO was obsessed with hiring people - to show huge amounts of growth.
I was sitting there saying "with a little bit of thought we can automate this stuff, go down the pub and watch the money roll in". It lasted about 2 years until everyone lost their jobs and everyone (VCs included) lost their noney.
Reminds me of someone bitching about spending a decade developing a solid business and then a bunch of VC funded companies came like locusts, eating everything, and then promptly died.
This happens quite a lot. There are some advantages to be had here if you play it clever but you better have a very solid play on how you plan on staying in business while the VC funded companies burn through their capital. At least you'll have a few opportunities to buy cheap Aeron chairs ;)
A camera drone that follows you isn't a bad idea necessarily. It can actually produce some amazing video for action sports athletes. But of course that will only be a niche market.
Why not apply this to vehicles? What if every car had a drone on the roof. Drivers could use the drone to look ahead and see how far ahead traffic goes, to see if that next exit really is worth taking. Or drivers could have it constantly surveying their vehicle, so that if a collision occurs, you have a bird's eye view of the incident, which is usually helpful.
Waze already answers if that exit is really worth taking in a much simpler way. Other use cases.. maybe.. but i'm not that excited that a drone follows most cars :)
If a drone could bring your electrical vehicle some extra charge by docking on the roof and then going back to closest charger... well that would be a true value added feature.
What do you think about this startup: Lucid Drone (https://www.luciddronetech.com/)? They use drones for cleaning houses/buildings. It's a part of YC S19 Batch 2.
Interesting idea, but ultimately drones are not strong enough to do what they want. It would need to carry itself, the tube, and counteract the force from the spray. Flight time is probably going to be way low.
While you guys were sitting around coming up with your own arbitrary units of measurement, we were actually inventing the two biggest reasons why America is still on US Customary units: the milling machine (in 1816) and the profile lathe (in 1818). (Two other important inventions from America that also created momentum for our system was the steam engine in 1825 and the platform scale in 1830, but I won't go further into those.)
Because the milling machine and the profile lathe were invented in America, they were built around the American measuring system, and accounted for America getting a head start in the Industrial Revolution. Those two inventions enabled rapid and consistent manufacturing of innumerable other inventions and tools at the time. By the time the Treaty of the Metre was adopted in 1875, America's manufacturing infrastructure was already 50 years old.
The benefits offered by a decimal-based system of measurement were not and still are not significant enough to justify throwing out an entire interconnected structure of tooling, tools, and machines.
I work for a custom fabricator (we focus on making highly customized one-off products that are either prototypes or highly specialized), so I have some insight into what it would take for just our shop to convert.
We have milling machines that date back to the 1950s. They still work really well and hold tolerances as well as any similar mill on the market today. We would have to replace them and all the tooling that goes with them. Actually, we couldn't just get rid of them. We'd have to keep them and also double (at least) the size of our facilities to hold the new equipment. Why? Because we have long-term support contracts with many of our customers. That means we'd have to have two sets of equipment and two sets of tooling. Some of our contracts are more than 40 years old, so we would have to plan on having both sets of equipment for at least another 40 years.
Also, we'd have to reprogram all our other CNC machines to work on metric units only. We also have to convert all of our drawings to metric, and many of those conversions aren't elegant. Try reading 38.5 millimeters on a measuring tape. Try marking it accurately and consistently.
Now, I also have to consider that my employees now need to know two sets of measuring units, and be able to switch between them, because there's going to be a long period where some customer drawings are in US units and some are in Metric. That means that sometimes they'll make mistakes, because humans are very good at switching. So now I'll have more re-work and scrap.
In short, we may be "insane", but while the SI countries were trying to figure out how to measure things, we were busy inventing and using the machines we needed to actually usher in the Industrial Age. By the time the metric system was finalized, we already had a 50-year head start on building things. Too late to switch. It will never make financial sense to change.
Attributing the success of US economy during the 19th century to the Imperial system is obviously absurd. Standardization needed as a bedrock for progress to be made -- the US simply inherited the British system while European scientists finally converged on a more logical system.
The truth is that you likely have to deal with metric units in your machine shop every now and then, just as we deal with imperial units in some places (in my country, pipes and fittings are often in inches even in new residences.)
Furthermore, the lathe has been around since at least Egyptian times, and the modern industrial lathe is thought to have been invented in France or the UK.
I also want to mention that I didn't attribute the success of the US economy during the 19th century to the US Customary system (which is different than the Imperial System, btw). I said that the standard that was in place in America when the Industrial Revolution began was the US Customary system because the metric system was still in its infancy at the time and not widespread. By the time the metric system was established, widespread milling, machining, and lathing in the United States was all built around the US Customary system. There was already too much infrastructure to throw out over a marginal improvement. It's a simple economic problem. Does the cost of throwing out all the old machinery, equipment, fasteners, drawings, and tools justify the marginal benefits of switching to a decimal-based measurement system? So far, the market's answer has been no.
I guess that's the first mover disadvantage. You get the tech first. It's not as good as second gen but you are sort of stuck with it since you already have a ton of infrastructure built around it. When the rest of the world adopts second gen and builds it's infrastructure around second gen and never bothers with first gen you lag behind hoping you'll catch up on third gen or so. The problem with units is that there's no third gen. SI system is just math and physics for the first and last time figured out correctly and that doesn't usually change for thousands of years.
In SI world there are still small pockets of old units. Like pipes measured 3/8, 3/4, 1/2, 1. The thing is they are isolated enough that almost nobody cares what the number means. Everybody just know that you need a 3/4 to fit with 3/4 you have. It becomes just a label.
> Two other important inventions from America that also created momentum for our system was the steam engine in 1825 and the platform scale in 1830, but I won't go further into those
Please do, I’d love to learn more about the American invention of the steam engine in 1825.
You could add drone-like control mechanisms to a balloon (or a balloon to a drone -- same difference, I guess), but that adds complications, so doesn't seem worth it.
This is true. The neat thing about the drone is that you can deploy it super rapidly though. I haven't seen it done but I am imagining 30 seconds or so.
Tethered drones are quite interesting. I fly one for our company but we have yet to get much interest in getting it adopted by first responders or LEO.
The "unlimited power" comes with some major asterisk from my experience although it is with a different system.
* Tethered power works great when plugged into a stable AC source, generators can be less than ideal.
* Weather conditions limit your altitude even more so, higher wind requires more power to stay up.
* The tether often works against a lot of the stabilization systems built into the drones, so flying them can be a handful under certain conditions
* The tether can get bound up on the spool
* The higher you go the heavier the tether is
Overall though, it's awesome technology with interesting potential. I have kept ours in the air for over 3 hours continuously and only landed so I could take a break.
Offshore racing sailboats use untethered drones frequently as there is rarely a helicopter available for publicity and action photoshots when 1000km out. I would think tethered would be great for them as they don't want to risk loosing a moneymaker with lots of potential remaining.
True, but in windy conditions, losing or leaving it behind might be a more likely scenario. Being able to retrieve it by pressing as button on a windlass or remote, then being free to attend to other things would be valuable.
Drones are huge in my department at school for collecting data with lidars and multispectral cameras. New drone techs for VC firms to invest in are going to need better technology, but a huge rollout of drones being used by everyday companies to do this is still in it's infancy. It's a good time to be someone graduating with experience with UAVs.
The numbers of uses are too large to count. I'll give one example: The DOT in my state has given large grants to use lidar from UAVs to be able to automatically identify wetlands.
Much of the drone industry circa 2009 - 2013 suffered from the exact same symptoms...develop awesome tech and then find a use case. I know of 3 that ran out of money before finding a way to even make money. sad but great lessons for anyone involved.
DJI is still valued at around $15 billion, right? To me it seems less like a drone bubble burst, and more like a single winner has emerged from round one.
They just got into FPV and I really hope they fail with their lock-in DRM, for the sake of keeping the maker spirit and FOSS innovation around in that field...
they used to have an open platform so you could bring your own flight controller. the most recent gen of their platforms locks you into their flight controllers (and their shitty api).
Heh. I have seen someone doing incredible maneuvers with crazy accelerations using a homebuilt drone. Compared to commercially available drones it's like an F1 car versus a tractor.
I think the drone industry is to a drone producer what the cloud is to hard drive manufacturers. Many of the startups were creating value adds atop DJI.
It was hyped in the 80s, and real advances were made, and then it crashed. Now it's hyped again, and real advances are being made, but it's still set to crash again. The problem is that "AI" is too big and romanticized a term. Symbolic reasoning is a real technology; neural nets are a real technology; AI is at best science fiction, at worst marketing fluff. So despite having some real substance, the hype still outpaces the reality.
Interesting that they did not mention 3DR. https://3dr.com/ They went from selling the hobby 3DR Solo (which I own and love) to focusing entirely on the government side.
The folks at Skydio are building something different. It's the first autonomous thing (car, boat, plane, drone, etc) that's widely available to consumers.
They're releasing a new product soon and if it's better than what they had before DJI or anyone else won't be remotely close to having a similar product.
Maybe a lot of the weaker companies have folded, hardware is hard, but would be very sad to see the innovation happening there collapse at this point.
This is a typical startup contraction. Remember when the "PC bubble burst" and dozens of companies went bust, all with similar but slightly different business plans. It turned out that being able to execute efficiently was the winning strategy, not features, not fancy graphics, not custom cases. Just ruthless efficiency.
A couple of years ago I went to UAV expo in Las Vegas and even then it was clear that a lot of drone companies were technology looking for a use case. And there were demonstrations of drones that did things much better (like mapping a traffic accident scene) than existing technology that saved both time and got the highways re-opened faster.
Racing them with PoV cameras continues to be a lot of fun, it just isn't the multi-billion dollar sport yet, give it time.
Except that that's already failed; it's used as an investment product, a get-rich-quick scheme, dealing in illegal products and services, untraceable exchange of currency (e.g. ransomware) and money laundering. Nobody pays with crypto. Except people that made a lot of money off of crypto and have a vested interest in its success.
If you look at transaction number and volume it has been basically constantly rising, so I'm not really sure what you're on about. "Haven't replaced all other currencies in all use cases" is not the same thing as "has failed".
Some 2 years ago quite a few services and shops were trying out bitcoin and offering it as an actual option for real-world payments. This trend has reversed, new offers are very rare and many companies who offered it in 2017 don't support it any more, as for any 'normal' business the amount of Bitcoin transactions they got (wherever I've talked with them) was extremely low after the initial buzz at launch.
The current number of transactions is not constantly rising. The current short term trend (https://bitinfocharts.com/comparison/bitcoin-transactions.ht...) is downwards. It's now about the same or less than it was in peaks of 2017 and early 2018, and whatever the growth in 2018-early 2019 was made of, it doesn't seem to be in consumer purchases of goods and services (i.e. being used as a currency).
Is it, though? I'm not saying I buy a lot of street drugs, but I know some people who know some people who do, and have literally never seen or heard mention of Bitcoin for this.
You know what I have seen being used, that surprised the hell out of me at first? Venmo.
Turns out people buy their friends a lot of sushi.
I've heard that anecdotally too, and tbh I would never ingest something a stranger sent me in the mail. I'm sure it's mostly fine, but there's a level of trust that whisper networks provide that is simply absent in an online setting. Heck, bitcoin itself purports to do exchanges in a trustless setting; trustless is the last environment you want to have for things you're putting in your body.
I've been buying drugs on the street and online for many years and online crushes IRL, just like it does for legal commerce. Online reviews have helped me always get top notch quality and exactly the product I expect. The street can't come close.
Except gold has a real industrial value (electronics), and is an exponentially more stable and liquid asset than bitcoin (albeit not as stable as goldbug scamsters claim)
Apparently nobody cares because only a small fraction of gold is used for that purpose. Most is used to sit on it.
Stable... hard to tell, let's see after a millenium or two. Gold had a pretty good run till now, but hauling one asteroid towards earth or figuring out how to cheaply filter it out of ocean water will drop its value to dirt cheap.
More liquid... probably. Market for gold is still larger for now. Let's check that after a decade or two. Trends don't look too optimistic for gold when compared to btc.
The FAA's licensing requirements for even toy craft killed this sector.
Good news though, it's going incredibly strong overseas such as in Yemen where they've bought Chinese made drones and rapidly upgraded them into a domestic high tech military drone manufacturing capability... with 1000 mile ranges and bomb carrying capability.
So drone fans might not be able to do their thing in the US, and the US might thus not have anyone locally that has a passion for drones, or is able to advance US drone technology, but the tech advances ... elsewhere.
Do you have any links that talk about this in more detail? Modifying a off-the-shelf drone (assuming you mean quad copter here, not a fixed wing craft) to have a 1000 mile range seems like quite the feat...
I traveled to the US in 2014 on a sponsored trip to research drones for agriculture. My report covered off on some of the technical challenges rather than the business side of it. You could see even in 2014 everyone had great ideas but profitable execution was going to be so hard. My report is here for a look back in time http://www.nuffieldinternational.org/rep_pdf/1438074605BenBo... . I then went on to start a business processing and providing satellite imagery in agriculture because I couldn't figure out a business model that would work for drones.
I have a first-gen Parrot AR.Drone collecting dust. Great fun initially until I started reading about flyaways and then got too worried to actually fly the thing outside. Even nowadays I get the impression this issue was never solved. At the time it seemed like one of those widely-reported but never admitted-to things. I never understood why the software wouldn't just go into a fixed landing routine if the control signal was lost or if a watchdog tripped. Still really curious about that actually, if anyone here knows more technical details.
So yeah that and the battery was never good enough. Now I look at drones and it seems like you have either really really cheap, subpar options or incredibly expensive DJI-quality drones. Something that costs around $300.00 AUD and that could fly for half an hour at a time would be a great toy, but I get that the tech isn't there. So I end up feeling like they're great tech that's ultimately killed by a thousand little tradeoffs. At least for me; if I was into video then one would be great for some things. That's a limited market in the end though, I think.
I always assumed it was just pilot error that the pilot blamed on the drone. Certainly saw enough of those situations that it's more believable than a phantom software issue.
My guess is fully autonomous drone products will prevail. The capabilities today are already incredible and the mobility lends itself to generalize to solutions that weren't previously possible. They could be one of the first kind of robotic agents that we regularly interact with in our lives besides automobiles.
Battery life is certainly an issue but even before a major breakthrough happens 10-20 minutes is a long enough flight time for a lot of autonomous tasks. Plus what if they dock and charge themselves?
I think they're going to continue getting smaller, more power efficient, and cheaper and it will continue to open up possibilities.
It kept getting harder and harder for me to legally use my drone. I own a parrot drone that weighs 500g. That uses to be the upper limit to not need an aviation license or certificate of some kind, and Insurance. Now 250g and up have those requirements, so I just stopped using it. I only used it once every few months for some arial footage anyway.
184 comments
[ 3.2 ms ] story [ 236 ms ] thread* Federal regulation of the aircraft has been slow to catch up, and is holding back many businesses from expanding.
* Non-drone companies across industry pull their unmanned aerial operations in-house
In short: there is no drone bubble. There is a short-term investment bubble.
https://news.ycombinator.com/newsguidelines.html
Subtantiating that CNN are pretty bad on this front.
It seems to me that if anyone should be pouring scorn on unsubstantiated techno-hype it's all of us here. People doing startups have the most to lose when snake-oil salesmen have their way and dominate the discussion. When the market then implodes we're left without funding and support for real technology that makes the word better. We get the reputational hit for BS we were no part of. Forgive us if those of us who are serious about technology and reality feel just a little emotional about that.
It needs to be called out by the knowledgable every time.
Especially when everyone who posts regularly here knows it to be true. Hype for snake-oil-tech is poison for us as a community, if indeed that is what we are.
"For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled." --Richard Feynman
You've probably seen that before too, or at least should have. Still worth repeating, often, in hype bubles of which there is clearly one surrounding drone delivery. At least I'm not trying to do something genuine, honest and real in that space. Must make things borderline impossible if you are.
Sure, but the way to do that is by sharing what one knows.
And quite a few tried pushing that a little too long an ended up with crashed drones. So my half cent is that batteries need to get better before people like me even consider them. Mentioned 0 times in that article.
Alternatively, drones need to get efficient enough to be solar powered, which is maybe the intent behind the acquisitions by PrecisionHawk Inc.
https://www.precisionhawk.com/blog/media/topic/precisionhawk...
More or less, they said they had given up on putting money into drone R&D until battery technology caught up, as they had plenty of ideas that they believed in but were unfeasible due to battery life.
Given the line of sight limitation for drones, gas power does not add much value. Batteries are usually sufficient to map/photograph the area within range of the operator. Given this legal limitation, it is often more economical to hire a small plane like a Cessna with an avionics package and a 400 mi range.
They exist, but they are expensive. https://www.youtube.com/watch?v=wS0Hr2lw604
You're looking at a fixed wing aircraft though with one of those. What's made the quadcopter drone possible is the high torque and instantaneous throttle response of small electric motors, combined with fast control electronics. The conventional way to fly a small helicopter is just like a full sized helicopter, with a main rotor and a mechanical swash plate.
With the right control electronics, these can easily be made into drones, but people tend to not like them as much anymore because they're mechanically complex and finicky. The electric motors are simple, and when one goes bad you just drop in a new one.
I think that’s a more impossible goal than increasing power density in batteries is.
This guy tries to make gasoline powered generator for large drone.
https://www.hobbytown.com/gas-rc-airplanes/c7574
This is hard to do with petrol engines, without costing a lot of extra weight or mechanical complexity, in a system where any single failure means a crash
There's differential based continous variable transmission that varies according to resistive torque on a control shaft: https://makezine.com/2012/08/10/lego-continuously-variable-t...
This might enable electronic control witout using any motors just one simple coil per propeller. Energy generated through this way of control could be stored to power onboard devices.
I don't really want to take it out into nature environments. That seems obnoxious where it's even allowed.
I suppose I could find clubs that fiddle with the things but I don't really have the interest.
So, basically, I have trouble imagining how I'd use them for more than a couple days.
If you're in an open area. And if there is only one drone.
A couple of years ago I was in a serene canyon in Utah enjoying the peace and watching nature when three drones decided to scare away the bighorn sheep.
People don't hate drones because they hate drones. That's why many seem to like the idea of Amazon/UPS drone deliveries. They hate drones because of what the drone pilots do with them.
I don't mean to be a dick; I've been flying R/C for ~25 years and I know how difficult it can be to find fun places to fly and how frustrating when other people get in the way of that. But I also really appreciate quietness and because I live in a city it's a rare gift for me when I do get out into nature. Please have some consideration for others and don't unilaterally decide what noise level is acceptable for them unless there are already significant sources of noise present or you're really sure nobody else is around.
Really very loud then for the people around you in a nature environment.
Last time I heard a drone fly overhead, I definitely felt anoyed. People tend to overlook the impact their toys have on other individuals
This just started in the last year or so, where virtually every time there's a drone off in the distance.
I don't have great hearing either (especially in one ear), after playing some musical instruments that you hold close to one ear to play. Those things are just loud and stand out when the other noises are the wind in the trees, birds, and the rustle of small animals in the underbrush.
If people can spend thousands of hours playing flight simulators on the computer, surely there's a lot of replay value of doing the same thing with an RC quadcopter through a forest or around your home/neighborhood w/FPV.
I havent bought one yet, but have flown a cool DJI/FPV setup around a local mountain. It was really fun
I bought a Taranis QX7 remote controller (~$100), which is amazing and I love it and you can use it with any other RC vehicle you want. I even hook up the receiver to my hardware projects so I can control them remotely.
You also need an FrSky XSR or R-XSR receiver (~$10), a Beitian BN-220 GPS (~$10), a motor and ESC (be careful, get a plane ESC, not a quad ESC, the plane one is larger and has better heat dissipation, ~$30). You need two small servos for the elevons, those cost about a dollar each. You also need a good soldering iron (the TS100 is $35 and is amazing.
For the controller, the Kakute F7 is fantastic and costs $45ish. It can fly in modes ranging from manual to completely autonomous (you get a Google Maps interface and mark where you want it to go and it just flies to those points). This may be a bit overkill, as it's the best board around, you can get a cheaper one (or maybe one that is made specifically for planes) no problem, but it's only $10 cheaper.
The FPV goggles can cost around $30 for some cheap ones, adjust those to taste.
I think that's it for components, although I should probably do a proper writeup if people are interested. There's much info out there already, but it might not be concentrated in one place. Let me know if you have any more questions!
Photography drone pilots will do all kinds of stupid shit that horrifies me, like fly near people, near obstacles, under trees, in crowded places, etc etc.
I blame it on the fact that a Mavic is super easy to get and requires no knowledge, whereas it took me two hours of reading just to figure out how to start my racing drone.
I’ll stick to writing code.
Fun enough for a few hours punctuated by recharging but not really something I see myself spending $500-$1000 on. I don't have kids. I can see them being fun toys. I spent many hours as a child playing with balsa wood gliders for example.
The main two types of flier before quadcopters were fixed-wing model airplanes, and helicopters. They are both riskier to use than quadcopters: fixed wings can't be stationary and need a large area to fly around, while helicopters have large dangerous blades. This is why either of these are less attractive to the mass market.
On the flip-side, I think converting a quadcopter to use an engine is expensive because controlling 4 blades at once is not as simple as with 4 electric motors. Either they need to have 4 gas engines (expensive, not as responsive as electrics) or they need to have a variable pitch mechanism (now you need an intricate/expensive mechanical system of levers and servos to transfer the control).
It's a scaled up version of his stingray.
The issue is that stability and control is easier to do with 4 blades rather than with one that has a swash plate.
The only cases where needing more time would be a big deal would be for filming something that lasts longer than the battery.
They are noisy, however. Drone buzz now is simply from spinning fans. Adding a reciprocating motor to the mix is all the louder.
The alternative is lighter airframes and solar power. For quads / octos, probably not viable, but for a prop-driven fixed-wing craft, a very small amount of thrust to counter gravity could be quite effective. Expanded polystyrene or similarly light body, solar cells over all horizontal surfaces, and an ultra-lightweighted controls and imaging payload. That would allow for long-duration flights at least in daylight.
The low-energy use would also make battery-powered night flights more viable, certainly beyond 15 minutes.
The principle cost is loss of hover capabilities.
also I believe if you want to be pedantic, if it runs on combustion it's a engine. if electric then a motor.
I think that might be less general than you think: General Motors hasn't exactly been known for their electric motors the last few decades.
Also see: https://en.wikipedia.org/wiki/Engine
https://en.wikipedia.org/wiki/Model_rocket_motor_classificat...
However "reciprocating engine" has a clear usage preponderance over "reciprocating motor":
https://books.google.com/ngrams/graph?content=reciprocating%...
They exist and are used for long distance drones.
Now, that's for small scale drones flying 15 minutes. For, say, commercial airplanes that can fly 18 hours, energy density becomes a major factor.
Build: https://www.youtube.com/watch?v=1yB49G756OI
Generator tests: https://www.youtube.com/watch?v=7zMPodObo-Y and https://www.youtube.com/watch?v=uf78udpz3aI
Not sure how well it'll work in practice but fun build.
I’m sure I would love them if I tried them, but I don’t want to be seen as the type of guy who owns a drone (like how I don’t want to be known as the guy who uses Linux or the village Libertarian, even though those words might describe me reasonably well).
There are occasionally bursts of genius in the field though. A few years ago I met a guy (Sergei at Fotokite, I believe), who was developing a tethered drone for photojournalism. Solves many problems at once: you don't run out of battery, you could control the drone via a stick or just pulling it around by the tether, people don't hate the drones as much when they can see the operator. He demonstrated it by deploying it in the reception area of an office building, probably 15x15ft. Now apparently they're developing a tethered drone for first responders that automatically deploys and gives you a 150' vantage point for situational awareness.
BTW: No relation to Sergei or Fotokite, just impressed with their tech.
You’d be dependent on the capacity at the exchange, might that not need to be upgraded too?
I was sitting there saying "with a little bit of thought we can automate this stuff, go down the pub and watch the money roll in". It lasted about 2 years until everyone lost their jobs and everyone (VCs included) lost their noney.
/s
There's probably other use cases too.
If a drone could bring your electrical vehicle some extra charge by docking on the roof and then going back to closest charger... well that would be a true value added feature.
Because the milling machine and the profile lathe were invented in America, they were built around the American measuring system, and accounted for America getting a head start in the Industrial Revolution. Those two inventions enabled rapid and consistent manufacturing of innumerable other inventions and tools at the time. By the time the Treaty of the Metre was adopted in 1875, America's manufacturing infrastructure was already 50 years old.
The benefits offered by a decimal-based system of measurement were not and still are not significant enough to justify throwing out an entire interconnected structure of tooling, tools, and machines.
I work for a custom fabricator (we focus on making highly customized one-off products that are either prototypes or highly specialized), so I have some insight into what it would take for just our shop to convert.
We have milling machines that date back to the 1950s. They still work really well and hold tolerances as well as any similar mill on the market today. We would have to replace them and all the tooling that goes with them. Actually, we couldn't just get rid of them. We'd have to keep them and also double (at least) the size of our facilities to hold the new equipment. Why? Because we have long-term support contracts with many of our customers. That means we'd have to have two sets of equipment and two sets of tooling. Some of our contracts are more than 40 years old, so we would have to plan on having both sets of equipment for at least another 40 years.
Also, we'd have to reprogram all our other CNC machines to work on metric units only. We also have to convert all of our drawings to metric, and many of those conversions aren't elegant. Try reading 38.5 millimeters on a measuring tape. Try marking it accurately and consistently.
Now, I also have to consider that my employees now need to know two sets of measuring units, and be able to switch between them, because there's going to be a long period where some customer drawings are in US units and some are in Metric. That means that sometimes they'll make mistakes, because humans are very good at switching. So now I'll have more re-work and scrap.
In short, we may be "insane", but while the SI countries were trying to figure out how to measure things, we were busy inventing and using the machines we needed to actually usher in the Industrial Age. By the time the metric system was finalized, we already had a 50-year head start on building things. Too late to switch. It will never make financial sense to change.
The truth is that you likely have to deal with metric units in your machine shop every now and then, just as we deal with imperial units in some places (in my country, pipes and fittings are often in inches even in new residences.)
Furthermore, the lathe has been around since at least Egyptian times, and the modern industrial lathe is thought to have been invented in France or the UK.
In SI world there are still small pockets of old units. Like pipes measured 3/8, 3/4, 1/2, 1. The thing is they are isolated enough that almost nobody cares what the number means. Everybody just know that you need a 3/4 to fit with 3/4 you have. It becomes just a label.
Please do, I’d love to learn more about the American invention of the steam engine in 1825.
You could add drone-like control mechanisms to a balloon (or a balloon to a drone -- same difference, I guess), but that adds complications, so doesn't seem worth it.
The "unlimited power" comes with some major asterisk from my experience although it is with a different system.
* Tethered power works great when plugged into a stable AC source, generators can be less than ideal. * Weather conditions limit your altitude even more so, higher wind requires more power to stay up. * The tether often works against a lot of the stabilization systems built into the drones, so flying them can be a handful under certain conditions * The tether can get bound up on the spool * The higher you go the heavier the tether is
Overall though, it's awesome technology with interesting potential. I have kept ours in the air for over 3 hours continuously and only landed so I could take a break.
What's the use case for this data?
Just waiting for the same to happen with AI.
Previous times the hardware was not there, now it is.
They're releasing a new product soon and if it's better than what they had before DJI or anyone else won't be remotely close to having a similar product.
Maybe a lot of the weaker companies have folded, hardware is hard, but would be very sad to see the innovation happening there collapse at this point.
A couple of years ago I went to UAV expo in Las Vegas and even then it was clear that a lot of drone companies were technology looking for a use case. And there were demonstrations of drones that did things much better (like mapping a traffic accident scene) than existing technology that saved both time and got the highways re-opened faster.
Racing them with PoV cameras continues to be a lot of fun, it just isn't the multi-billion dollar sport yet, give it time.
Did someone say blockchain?
I wonder what percentage of those transactions are generated by people trading in cryptocurrencies vs them actually being used as a 'currency'.
The current number of transactions is not constantly rising. The current short term trend (https://bitinfocharts.com/comparison/bitcoin-transactions.ht...) is downwards. It's now about the same or less than it was in peaks of 2017 and early 2018, and whatever the growth in 2018-early 2019 was made of, it doesn't seem to be in consumer purchases of goods and services (i.e. being used as a currency).
This actually is a currency use case. It's an illegal one, but it's a really useful one.
It's not clear to me that it's a use case that requires a blockchain, but i don't think there are currently any viable competitors.
You know what I have seen being used, that surprised the hell out of me at first? Venmo.
Turns out people buy their friends a lot of sushi.
Apparently nobody cares because only a small fraction of gold is used for that purpose. Most is used to sit on it.
Stable... hard to tell, let's see after a millenium or two. Gold had a pretty good run till now, but hauling one asteroid towards earth or figuring out how to cheaply filter it out of ocean water will drop its value to dirt cheap.
More liquid... probably. Market for gold is still larger for now. Let's check that after a decade or two. Trends don't look too optimistic for gold when compared to btc.
It's better for electronics than copper, so there's that.
> but hauling one asteroid towards earth
Seems impractical, expensive, and the yield isn't clear.
> or figuring out how to cheaply filter it out of ocean water will drop its value to dirt cheap.
You should check your math on this.
342,600,000 pounds of gold have been mined[1], but the oceans hold 20,000,000 pounds[2].
1: https://www.bbc.com/news/magazine-21969100 2: https://www.atlasobscura.com/articles/gold-ocean-sea-hoax-sc...
Good news though, it's going incredibly strong overseas such as in Yemen where they've bought Chinese made drones and rapidly upgraded them into a domestic high tech military drone manufacturing capability... with 1000 mile ranges and bomb carrying capability.
So drone fans might not be able to do their thing in the US, and the US might thus not have anyone locally that has a passion for drones, or is able to advance US drone technology, but the tech advances ... elsewhere.
So yeah that and the battery was never good enough. Now I look at drones and it seems like you have either really really cheap, subpar options or incredibly expensive DJI-quality drones. Something that costs around $300.00 AUD and that could fly for half an hour at a time would be a great toy, but I get that the tech isn't there. So I end up feeling like they're great tech that's ultimately killed by a thousand little tradeoffs. At least for me; if I was into video then one would be great for some things. That's a limited market in the end though, I think.
I always assumed it was just pilot error that the pilot blamed on the drone. Certainly saw enough of those situations that it's more believable than a phantom software issue.
Skydio - https://www.theverge.com/2018/4/5/17195622/skydio-r1-review-...
Shield AI - https://www.shield.ai/nova
Battery life is certainly an issue but even before a major breakthrough happens 10-20 minutes is a long enough flight time for a lot of autonomous tasks. Plus what if they dock and charge themselves?
I think they're going to continue getting smaller, more power efficient, and cheaper and it will continue to open up possibilities.
They do B2B drones though, maybe this article was too focused on B2C companies.