DJI seems to be the one company that requires sideloading the Android app. DJI makes great products, but there's really few companies who I would trust less to give unvetted access to my phone.
I know the Android app sandboxing should help a lot, but there's enough unknowns & potentially lurking security holes that I really wouldn't want to trust them. They also are 100% the most likely company on the planet to self-update & start truly-epic-level spying on the world if China ever gets a little bit more nervous. It's just not a tenable proposition to trust DJI with this security stance; no one at the company has the power nor right to say no, & as such there's no basis for trust to form.
I get why people want the end of the App Store monopoly and I apprieciate the idea of being able to run anything I want on my hardware, but this here is the reason why I'm hesitant.
As a user I want the big developers to have to bend the knee to Apple. I don't want to side load my DJI app, I don't want to download Instagram from the Meta store.
I don't agree with all of Apple's rules (and also with the way the app revenue is split), but I'm mostly satisfied with them.
Then the future of computing will be in the hands of mega-corporations. Have an innovative app idea that goes against corporate policy? Too bad, won't happen.
Strict sandboxing and a robust permissions model can still give you security with these kinds of apps.
It really bothers me that DJI is basically the one notable example that gives the anti-sideloading FUD-dites some ammunition. It's generally not a real issue, DJI are just extra punks, and yeah, sandboxing should offer really good protection in general, but maybe maybe your drone is live uploading all drone data to the CCP perhaps.
Freedom enables choice. Humanity's greatness is built entirely off our ability to explore possibilities & create new ways. Gating innovation behind what Apple & all the governments permit us is just not an option for the most important most noospheric device on the planet. Not having choices is to let ourselves be chained up in the PKD Black Iron Prison. I refuse that dejection.
What happens when Android starts clamping down, as they have? Android ever more ratchets shit down, and more and more essential apps like Termux simply can't make it through the Google gauntlet. What happens if Android stops allowing sideloading?
I hear this defense a lot from the anti-sideloaders, that it's fine that this platform have no freedom, because there's another popular platform with some freedom. Even with the DJI example here before us (which so far has been icky but had no known actual bad consequences), I roundly do not get the argument, do not see the appeal of being trapped, of arguing everyone else on this platform needs to be under the fist of Cook & every government on the planet & without feet to stand on.
And it's repugnant to me that the defense is that there is one other platform where freedom exists, as if it's OK that one set of folks fall down the slippery slide & leave the rest of humanity tottering for themselves, trying to keep feet under them against perpetual universal software helplessness.
Apple has basically been winning the AT&T position against Carterphone. They've used monopoly power to control push around dominate & extract, to make up whatever terms it wants for the world. That's a position that damned & binds humanity, and one we have rejected already quite clearly, but now that computers are involved it's free game to coral the humans & other entities & to rollback the Carterphone right to modify the world about you.
There's one plan I'm OK with here, which is Apple can keep it's closed platform but has to let other OS run on iPhone & make equal use of the hardware. If apple demands control over iOS, I think that's an alternative out for hardware owners.
>There's one plan I'm OK with here, which is Apple can keep it's closed platform but has to let other OS run on iPhone & make equal use of the hardware.
This is an interesting idea. It sounds pretty good actually.
It hasn’t yet, but if they do there are already alternatives in the Android world. That might the chance that Microsoft, Amazon, Epic, or any other company might take to launch a phone that allows sideloading.
Yeah punks totally, doing punk activities like profiting off the genocide of an ethnic minority. I absolutely want to give unfettered access to my device to an economic arm of the CCP and anyone who's against that is just spreading FUD
This is the third or fourth time I've heard this sentiment lately and I don't understand it. Why would the ability to sideload apps impact the ability of Apple to curate app quality in their app store?
Given that I can sideload Android apps, why can I still download Instagram from the Play store?
Well, that's easy. Developers who are big enough (like Meta or Tencent) can leave App Store and distribute the apps through their own store where they don't have to succumb to any rules. The customers will have to follow.
Now, you might want to say, I don't have to use their apps, and you are right. But I'm happy with the current situation. I want to use Instagram and I want Apple to negotiate what the app can and cannot do for me. I want to fly my DJI drone. (You cannot download the DJI app on the Play Store. You have to sideload it on Android. On iOS it is in the App Store because it has no choice.)
I’m in the same camp. I got my son a Mini 3, which is a stellar quadcopter in its own right, but the DJI ecosystem is a spectacle to behold for some good and many bad reasons.
One issue could be that they don’t communicate compatibility between devices very well, so it’s very hard to know if buying one thing will lead to being very limited and locked in — even within the DJI ecosystem. People often find themselves needing to buy more equipment to accomplish seemingly trivial things. For example, the controller we got can’t install apps and it doesn’t work with a multitude of other quads DJI sells. Why not? Are they good reasons or bad? Where can I find clear explanations? I’m sure the controller is very resource constrained, but can’t I just opt into bad performance if I don’t mind? Is the controller incompatible with other quads due to radio protocols, or is this a choice intended to force eventual upgrades?
Without being able to tell with certainty, sometimes it begins to seem a little manufactured and nefarious.
Otherwise, the fact that their protocols are all proprietary and I can’t use the controller with custom quads is very frustrating. I have a Radio Master TX12 that I can’t use with any DJI products, and a DJI controller I can’t use with any custom quads already built or that I could ever build.
That’s all crazy. Sure, tight integration means better performance, but the extent to which it’s closed off is unreal.
Although the Mini 3 is a genuinely awesome device, I don’t think I’ll buy DJI again. I think I’d rather buy into open VTX hardware and use my own camera to record footage. It’s a huge compromise in convenience and upfront cost, but if you care a lot about the hobby it’s perhaps well worth it. The flexibility afforded is enormous.
100%. I was thinking about getting an Avata, and I was just about to pull the trigger, then they released a recent update that includes Remote ID which means that now you have to connect your phone to the goggles to take off (something you never had to do before) so it can get a GPS location to broadcast the operator location. I'm not even sure why it would need to do that because the drone itself has a GPS so wouldn't it know where it took off from anyway?
I'm in Australia so we don't even use Remote ID, so I don't know what will happen if I bought one. Do I have to connect my phone to use the drone? Will I have to in the future? How can I find out? What if I'm in an area with no phone signal?
I am quite interested in trying out FPV drones, and I would like to have the (unique?) feature of the Avata where if I get in a bad spot I can just hit the switch to enable GPS auto-hover, but all of the uncertainty with DJI stuff has made me not buy one, and I probably won't.
I didn't even know you had to sideload the app until this thread. Spidey senses going off 2000%.
I don't need to spend $2k on another hacking project on my list of devices that I've paid for but still don't really own. Probably better to just buy a more DIY setup that I have full control over, and real ownership of.
GPS return to home/fail safe modes has been a feature of most fpv drone firmware for quite a while. Betaflight, iNav etc .. All support that as long as you have a GPS (+ compass) module installed.
The only thing avata brings is some beginner friendly flight modes for a premium. Kind of like Apple products.
You also have a lot of ready to fly/bind and fly options (buy a radio separately and bind it to your quad and fly) out there these days..
Are there DIY digital FPV goggles with as good quality video as the Avata ones? Do you need a separate camera for the footage (gopro), as well as the FPV camera?
Also the gimbal on the Avata seems very well integrated, even having fancy head tracking, which I wouldn't mind giving up, but do the DIY drones have a gimbal at all, or is it a manual job to stabilize the footage later?
I agree, DJI seems like the Apple of drones. And I'm against Apple for computers and phones, because I can DIY better and more features.
But for drones, I suspect that DJI actually has features that you can't (reasonably) DIY.
What are the good brands for the ready to fly/bind and fly drones, so I can have a google?
What quad you want to buy depends on what kind of flying you want to do. There are all kinds of them for freestyle, long range etc.. If you want something with prop guards, here is a decent one: https://www.youtube.com/watch?v=aFBBTtNZUPc . See if you like it? Geprc, iFlight are usually decent brands. When buying, go for ELRS 2.4 Ghz version with GPS. ELRS is just an open source rc link system that has really good range.
Gimbals and headtrackers are possible with DIY quads. But that is mostly used by fixed wing people. For just racing/freestyle kind of flying you don't want those kind of things. It is just one more thing that breaks. Here is an example of a DIY headtracker: https://www.youtube.com/watch?v=YVC5HyJGU4Y . I think DJI just has some beginner friendly features with additional sensors for collision detection, automatic takeoff/landing etc... Once you are used to flying, those are mostly useless.
If you want an alternative to DJI video system, there's Walksnail Avatar, which is quite decent too (Comparison video: https://youtu.be/d0VoBp9-sn4?t=550 ). Both DJI O3 Air Unit and Walksnail Avatar have onboard recording system which eliminates the need for a GoPro. At least for non/semi professional footage. For stabilizing the footage, people just use Gyroflow: https://gyroflow.xyz/ later on.
As always, before investing in Goggles/Quads, do get yourself a good transmitter and practise flying on a simulator. Maybe Radiomaster Zorro or Radiomaster Boxer: https://www.youtube.com/watch?v=b0b1tQ-hyCQ . ELRS version.
> Really not a fan of getting trapped in closed stuff with forced firmware updates that take away features.
Ahh in that case avoid DJI Goggles 2 and go for Walksnail/Dominator HD ( https://pyrodrone.com/products/pre-order-fat-shark-dominator... ) Goggles and their video transmission system. DJI requires you to register and update your system before use iirc.
If you are interested in a more open source system, there is also HDZero. Goggles software, CAD Files are all out in the open. Image quality is so so, because it prioritizes fixed latency. You get 3ms glass to glass latency (Compared to 28-30ms in DJI/Walksnail). Comparison between Goggles: https://www.youtube.com/watch?v=TMOeIQ4VRX4
Pretty sure this comes with it's own standalone DJI remote (which is really just running android underneath). No need to sideload or connect to your phone in any way.
- No need to deal with Google or the Play Store and everything that comes with, having to do reviews and all that
- Faster deployments to actual devices. AFAIK, you could even implement self-updating if you're feeling brave
- No sharing of data to Google and/or Play Store.
- Basically everything that the Play Store is now allowed, as you set your own rules (besides local regulation).
Basically, lots of goodies for developers and companies who want to skirt whatever safety Google has setup, with minor benefit to the users themselves.
Presumably, no benefits to users (other than being able to use app itself, that is). But I believe even sideloaded apps are scanned daily by Play Protect[1] which should spot if DJI started doing something sneaky?
When will these compact drones become less noisy? I know there are new propeller designs being researched to be much less like a swarm of bees is flying around near you, but don't know how far off they are to retail commercialization. It just makes it hard to stomach flying a drone in a pristine environment knowing you are disturbing all human and non-human animals nearby.
There's some interesting research around toroidal propellers which move the same amount of air, but create far less turbulence and so make a less annoying noise. However, they're not going to get significantly quieter until the energy density of the batteries gets significantly better (a lot of what you're moving with a drone is your energy storage).
There's always a hope that someone might invent a really silent propellor, and there's some funky stuff you can do there with trailing edge profiles, but I don't think we'll see a huge leap beyond toroidals for a while.
It does a similar sort of thing - asymmetric offset propellors shunt the frequency of what gets emitted way down. It still emits a similar amount of energy, but the direction of emission changes, and with the lower frequency, there's a lower perceived sound. Toroidal props actually have less sound coming off them, and at a lower frequency. I've got some ideas about how you'd potentially then have a greater spread of frequency by creating an offset toroidal, but I suspect the efficiency goes down, and I've got better things to do with my time than model sound profiles of weird, hard to manufacture (other than printing) prop geometries.
Related note - the offset bit is similar to how car tyres work by having patterns in the tyre which spread the frequency response generated by each, so they don't have a multiplicative effect.
(I can get really boring on applications of acoustic design but I'll cut off there)
To some degree it's inescapable with the small props. The higher end DJI models have enormous props which are better for noise as well as endurance. 16" props on the Inspire 3.
The natural extent of this progression is to have one or two very large 2-4 blade rotor (for low disc loading, high endurance, lower rotor speed and noise), with a small auxiliary rotor for yaw control if only using one lift rotor.
Yes, the natural endpoint is in fact just a helicopter. Only reason we don't see more of that (edit: in photo drones specifically) is the mechanical complexity.
I've not actually measured the volume of various props - I can, and should! - but based on my memory I don't think it's quite as simple as "smaller props make more noise".
For FPV stuff, I've flown everything from 25mm to 9" (prop size). The ones that seemed quietest were 2.5"-3", very light, and designed for performance.
I think it has to do with the pitch of the noise and the "idle point" of your motors.
Smaller props turn faster, and the faster they're turning the higher the pitch of the noise they produce. At some point it borders on inaudible. Even if it's technically producing more decibels (It might be! I've not tested that.), it isn't as disruptive. Larger props turn slower, producing less noise and at a lower pitch.
"Idle point" is what I'm calling how much throttle you have to give it to hover. For my favorite 5" freestyle FPV quad, that's 18% (+/- 2% depending on temperature and humidity). It's acceptably quiet if I'm just hovering or slowly moving around, but sounds like an angry weedeater if I'm flying aggressively. The idle point for my 2.5"-3" quads is probably around 25%; my 7" long range build is 12%.
The more throttle you give it, the faster the props spin, the louder and higher pitched it is.
I saw that! I’m sure they will patent it and consumers will be stuck with more traditional designs. Sigh, the patent law really needs a refresh, considering how rapid research and market response became.
Exactly. All this focus on improved image quality, battery life and whatnot, but no mention about noise levels, or even R&D into the area.
Currently, only the video is usable from drone footage, which is a huge limitation. All drone footage either has music or an audio track from an external source overlaid on top of it.
I wonder if the propeller and wind noise could be cancelled or removed with some fancy algorithms. It wouldn't eliminate the produced noise, but it would at least make the audio track usable.
I've used on-board audio a few times, but it's rough.
The only case that comes to mind where it was really successful IMO is if you're using an FPV quad to chase a vehicle: race cars, dirt bikes, monster trucks, etc. All of those things are loud enough that they're clearly audible over the prop and wind noise. If anything, for dynamic chase footage, the sound of the quad's motors spinning up adds interest.
Otherwise, you're not going to hear anything worthwhile from the air.
Are they really that noisy? I fly 10" drones and once they are about 30 ft away you can't really hear them. 3" and 5" drones are silent at 30ft on a calm day.
This strikes me as a rather callous lack of empathy. Personally, I find the mosquito whine of drones to be really annoying - especially when I am trying to enjoy the peace and serenity of a natural environment. There is nothing more obnoxious than going surfing only to have some creep buzz you with a drone. It's really intrusive and the sound is plain just obnoxious. I'm trying to enjoy the ocean like people have done for eons and someone else can't seem to enjoy it without flying a high pitched squealing camera over my head? It's a shame too because as a camera enthusiast I would love to be able to take shots like the ones DJI features on its website. They are spectacular views. But the obtrusive rudeness of the technology is the one thing holding me back.
There is a farm I regularly pass on my bicycle and closer to the end of the summer they have boatload of deer coming in an wandering on the field. I have zero clue what they do on that field. Anyways those pay zero attention to my drone and go about their business whatever that is.
I fly an Autel Evo 2 Pro, and love it - but I still recommend people new to drones pick up a used DJI Mavic Mini (or Mini2) when just getting into aerial photography. You can pick them up barely used, with three batteries, charger, and carrying case for ~$250 if you shop around.
I've had five of the things at this point. I keep buying them when I find them cheap, and passing them on to people who are entering the hobby/field. It's always a good idea to have a backup when I'm working a job, and the Mini series is small and light enough that I can throw it in my camera bag even when I'm not planning on flying.
Mavic Air is great but I had the opportunity to also flythe Mavic Air 2 in the same session and it's much better in regards of range and dealing with (heavy, urban) interference. If only I can only upgrade the radio/remote to the newer one..
I doubt they care much about you taking some videos on holiday! Probably don’t use it to fly over military sites and there is zero risk.
I worry that some of the worlds great companies get caught in the power wars of the US and China. Both countries working together has been so good for the world - nobody will be as prosperous without the other - but who cares as long as the powerful people in the US and China get to maintain or increase their power.
With the apps the issue isn’t the ability of China to harvest data on someone at a kids sports events or what not or what basketball team they like.
The issue is that for example with TikTok the American citizen/national with relatives has been watching videos on <sensitive topic> and now their relative back in China had a visit from the authorities, or this info gained from the app on this politician or journalists or their family could be used for leverage, or a number of scenarios like that.
The chinese authorities already illegally sets up “police” stations abroad in various countries. The authorities already do apply pressure to relatives still in China.
This is the beauty of a product like their drones that records detailed telemetry, and can upload video and images to their cloud service: you get loads, and loads of data that people with zero experience in intelligence gathering think is completely harmless. There really is no way of knowing how much sigint the drone is capable of gathering and sending back (it has an SDR chip).
Just in the area I live there are perhaps a dozen military and strategic civillian installations. This is going to be true for any city. You'd be surprised how much civil infrastructure is at least somewhat classified. Because compromising it can have serious consequences.
Some of these installations are even known to DJI and are geofenced off. In fact the DJI maps actually identify some military installations that are not identified on official maps provided by local providers. And both the maps and the geofences seem to be very actively maintained - much more so than the locally obtainable maps. But you can still get close enough to image a lot of these areas.
The fact that they have an SRD on board, and because there are probably on the order of a couple of thousand DJI drones in the city, AND they phone home, you could get drones to do targeted scans depending on what area a drone is in. I'm not saying that they do this - but I am saying that all the things China would need to leverage DJI drones as a global network of intelligence-gathering assets are already there. The safest assumption is that they do leverage the drones for intelligence gathering.
As for opportunities: Another good example is when there was a NATO exercise here last year. While down in the harbor to look at a US Marines ship (can't remember the name) that carries MV-22B Ospreys, I spotted what appeared to be a Mavic Pro of some description in the air. It probably wasn't the only drone trying to capture footage of the exercise.
So yeah, exploiting naive users to unwittingly take part in intelligence gathering is a pretty neat trick. And it is a bit surprising that the regulations for this are so lax given the immense potential these platforms have.
1) Other manufacturers don't have restrictive geofencing like DJI - it's not a regulatory issue, but liability.
2) DJI's flight restrictions are determined by DJI, not the FAA. They don't align with each other, so you end up having to keep track of and comply with two airspace systems.
That's why I abandoned DJI for my business use. I wasn't able to take off to cover an event at an airport for which I had FAA approval. DJI gave their approval... but the code didn't work, and it required that I have Internet access on the phone I was using to fly to boot.
DJI arguably makes the best consumer drones - and as a platform I love my Mini Pro 3. It's small and light enough to easily put in a pack, easy to fly, and just works well apart from DJI's geofencing system.
HOWEVER - I'm heavily restricted from using it at my own property, not because of the FAA or any other government restrictions, but because of DJI.
I live about 10-miles from an airport. Per the FAA's on "B4UFLY" app - my rural property is clear of any restrictions, outside of the altitude restricted zone of the airport. However, DJI draws their own "FlySafe Geo" map differently, and places me in an altitude zone where I'm limited to 150-meters.
This wouldn't be so bad, except that I live next to a large hill/small mountain, that's a great place to fly over and take photographs.
If I were to go to the top of the mountain (also in DJI's altitude restricted area) and take off, I could fly out 150m from the top of the mountain, no issues. But since DJI takes no note of the ground height changing as you fly, if I take off from my property and try to fly my drone to the top of the mountain, I'd crash into its side long before I got there.
There's really no solution to this issue, besides hoping that someone creates a hacked firmware for my drone that ignores DJI's restrictions, like they have for other DJI models.
If another manufacturer developed a comparable drone that didn't suffer from DJI's restrictions, I'd buy it in a second.
Autel's app will pop up a warning - a small warning that's easily dismissed and doesn't get in the way - if it thinks you're in restricted airspace, but won't prevent you from flying.
There's one Asian country - Japan? - where the Autel app does restrict based on geofences, but otherwise it's up to the operator to ensure regulatory compliance.
AFAIK in Norway (all of EU?) you need Internet in order to fly close (regulated, not closed, airspace) to an airport - the drone self-reports position and id to air traffic control, so they don't show up as unidentified blip on their radar.
In my home town basically the entire air space is regulated due to the location of the airport/flight paths.
My drone stayed without connection for more than a year already. Sure it politely says at startup that said zones need an update but other than that keeps functioning without a hitch.
I am way more concerned about domestic spying. I would not give a rat's ass if DJI steals my pics and locations. I photograph what I consider pretty places and that's the extent. They can have it.
Skydio seems to be doing just fine, though their customers are more companies that want security rather than consumers.
I think what killed it is the FCC rules, which make it illegal to fly in a large number of places where it really shouldn't have to be. This severely limits the size of the market for a $1,000 gadget.
DJI is a vertically-integrated beast with CCP ties; it won the monopolist position in 2015 against venture-backed 3D Robotics. Skydio, the largest US entity, really only exists as a protectionist hedge.
Consumer products mean low prices, the opposite of what America does well.
> Consumer products mean low prices, the opposite of what America does well.
This is actually sad. Ford could drive down the cost of automobile down to $4000 while its workers could earn $5 an hour. America used to be the industrial behemoth what could manufacture and innovate better, faster, and cheaper. What's interesting is that there was a powerful country that would mock the US for producing things cheap but not necessarily crafty in late 19th century and early 20th century. Yeah, that was Britain. We all know how history went.
we gutted most of our manufacturing and outsourced it to china decades ago. we Let our entire non military manufacturing industry die on the vine so there isn't much of a culture of supply chain left.
blame harvard business school. The cumulative damage of their Alum to the american economy is astronomical
Can we stop with this nonsense myth? The US is the second largest manufacturer in the world. Does manufacturing make up as large a share of our economy as it did in the past? No, of course not, that's a normal progression to a service economy, which is why the US is so wealthy.
> The US is the second largest manufacturer in the world
The US used to the largest. I was wondering if we should stay the largest, and most importantly the best. Case in point, the average age of workers in our shipyard was more than 55. The number of shipyards in the US has been decreasing over the years. I can see how that will help us innovate on and build better and cheaper ships, for military or not. Another data point, Intel can't compete with TSMC when it comes to manufacturing chips. Plus, if the so-called "low-end" manufacturing is gone, how could we possibly train the next generation of engineers? My assumption is engineering will be like writing software, a person who can only draw boxes will not be a good software engineer, and the person could not even learn draw boxes well without years of programming experience.
Meh, there's been fears of imports damaging the home country's manufacturing/economy for hundreds of years. It's why tariffs have existed for ages. If the government valued manufacturing at home, they've had the tools for years to make it happen.
Of course, are citizens ready to pay more than double in some cases? Not sure.
Without a booming civil manufacturing industry, how could our military industry stay competitive? There was a reason that a toilet for Airforce would cost north of $20K.
There's a whole synergistic Shenzen ecosystem of companies, and the US is a very expensive place. Most of the people with the skills to do this are already working for either FAANG or the long tail of other US tech companies, not on something as comparatively low value as drones. The US is simply higher up the value curve for manufacturing.
It's much more expensive to manufacture lots of stuff in the US, but clothing and electronics probably top the list. Most people aren't willing to pay 3x the cost of an equivalent item just for the warm-n-fuzzy "Made in USA" sticker on it. A few are, but most aren't.
From a hacking perspective, DJI previously seemed to be pretty nice with SDKs etc allowing you to get the video feed, send controls to the drone etc. But from what I can see they stopped supporting this for most (commercial) drones from Mavic 3 and onwards... Anybody know otherwise?
What I'd want is to be able to remote control it from any other device, it doesn't really matter if it's an Android/iOS app or running on linux/windows.
>We did not detect any leakage of data to DJI servers with any of the apps.
However, all three apps make a Domain Name System(DNS) request to the computer hosting
the WiFi network to resolve the IP address of "www.dji.com" when they are first started.
This DNS request goes to the router or a DNS server, and does not actually communicate
with any DJI servers, but it does tell the app the IP address of DJI servers.
The DJI Pilot Beta app follows this DNS request up with a ping to one ofthe DJI servers, and
receives a ping response back. This ping is used by the app to check that the server is up and
can receive messages. The ping does not contain any data, but the server could get the IP
address of the tablet and the time the app was started (since that’s when the ping is sent)
Note that the CCP can command a device to phone home in the future. So "We did not detect any leakage of data to DJI servers" just means that the device is currently not sending data home.
Honestly? Assume they collect everything, such as location, the pictures and videos taken (and not taken), flight paths (this one I know for sure), telemetry from the device and drone, etc. Even if they're not doing it now there's no reason they wouldn't in the future, they are basically a monopoly when it comes to consumer drones and they are in China, they don't give a ** about privacy.
Ironically, I do recommend them, they simply are the only option in the market right now and all of their drones have something interesting to offer.
If you have log sync turned on - which I think is the default - your entire flight path is sent to the cloud. There are also a number of thumbnails from the flight visible in the web app. They're very low resolution, but I don't know if that's all that's sent or if they're downscaling them server side.
Every device manufacturer tends to fib a bit on battery life vs. real world, however, 45 mins of flight time is impressive. Even if its 30-35 mins fully maxed out that can give you a lot of time to creat content and explore
I agree. I get fairly bad range in my small city, even with line of site. I get serious lag at around 2km. In places with less interference, it only bumps up to around 2.75–3km. It’s advertised as having 10–15km range or something.
On the other hand, 3 batteries in a Mini 3 is pretty much guaranteed to keep it in the air for 90 minutes. It’s very impressive. Our custom quad manages more like 20m per flight with twice the battery capacity (though, twice the weight as well).
Tangentially: I wonder if a really long-flying quad copter can be made using a regular ICE + generator as a source of electricity. Carbohydrates are still more energy-dense.
It has been done, you can buy Chinese made very light 2 stroke engines packaged with a generator. Used for large hexes and octocopters. Incredibly noisy and the lightweight design of the engine means it has a very short time in hours operating before rebuild.
These are a known thing from like three or four different vendors in the community of people who build medium to large sized ardupilot/arducopter based things.
A hybrid helicopter with 4 blades? There's a lower limit for weight to fit the ICE (we can't scale an ICE down much), so it might end up pretty close to a smallish helicopter
The trick with the tiny little model aircraft engines is that you typically get a decent lift/drag efficiency multiplier with fixed wing vs. quadrotor. In my experience, there seems to be a somewhat scale invariant power requirement of ~200W/kg (applies to a 249g Mavic Mini and much larger 25kg units). For a fixed-wing, you can get a 5:1 or 10:1 L/D and get away with more like 20-40W/kg instead.
You also probably don't want to direct-drive a multirotor off of the gas motor but rather convert it to electricity so that you maintain the rapid control authority on the motors. There are companies that make suitable powerplants, e.g. https://www.pegasusaero.ca/hybrid-advantage
I assume the issue is not the energy density but the support density.
I have a 3DR Solo and it can carry up to 1 lb of extra weight for 20 minutes or so on a full charge, and run for 30ish minutes while only carrying its gimbal.
The battery in the 3DR Solo weighs about 1.3 lbs itself, and the frame is by no means lightweight. Assembled it weighs at least 70% more than the Mavic 2 Pro that I occasionally get to use at work.
To fly my drone, the 3dr solo draws an average of 14.2 volts at 15 amps (so close to 200 watts on average) but up to 800 watts at peak.
You would need an ICE and Generator combo that can put out an average 200 watts of constant power but that can peak at 800 for short periods that weighs less than 2.3 lbs including fuel that runs for more than 20 minutes to equal that 1 battery.
For instance, this is one of the most well known and light RC ICE machines:
It weighs 9.5 oz and runs for 45 minutes on 6 oz of diesel (probably not at maximum output, but still.)
So you're at roughly 1 lb (assuming 1/2 oz of screws and tubes to connect the motor to the fuel and both to the frame.
However, there is an issue here. This motor is rated at .35 hp. 1 hp is roughly 746 watts of power. That means max rev, full bore, this 1 ICE only has enough power before conversion losses to hover my 3dr solo in calm winds.
This one is much better. 1.8 hp at max revs, so 1300+ watts. But, it weighs 1.87 lbs.
Add in fuel and you're at the max, so good to go, right?
Well, the next problem is that even though we have an engine, we don't have a generator to convert from fuel to electric.
You would also need some supercapacitors or batteries to store the charge, some system to integrate the control of the motor into the drone, more mounting materials.
And, you are already over the weight limit. That doesn't mean it can't be solved, for instance you could use a hexacopter or octocopter to increase how much carrying capacity you have.
Basically, you can do it, but there are better ways to do it. You would have to design an ICE RC Drone from scratch accounting for all of the intricacies and difficulties of using combustion engines in a drone and controlling them safely just to get what will probably be a slower, more expensive and more fragile drone that breaks down more often and needs frequent servicing intervals.
This is because software development still doesn't and has never qualified as capital-E Engineering. If "software engineering" were held to any of the same kind of rigor and responsibility that applies to mechanical, civil, aerospace or nearly any other "Engineering" discipline, we might not be in the technological & sociopolitical hellscape we are in right now.
Speaking of that software I resurrected an old DJI drone for a family member and after takeoff the app prompted me with so many warnings/errors/updates/notifications that I crashed. Truly horrifying to see a drone drifting towards you (and others) while you're furiously tapping to close warnings.
The active tracking is very impressive (see this review, 10 mins in): https://www.youtube.com/watch?v=ejmGSZtLu6I The way it flies sideways & dodges obstacles, even while the person is running all over the place, is awesome.
I was sailing in Croatia (on sails) and we’ve spotted some dolphins. So I launched a Mavic Air to observe them from closer. Soon the wind picked up and so did the waves. I piloted the drone back to the boat and planned to catch it in my hand, but I haven’t felt comfortable doing so, since the boat moved like crazy, while the drone was static in the air. Don’t forget to imagine the million ropes and wires in the air, which is just the reality of being on a sailboat. Wind howling, captain shouting orders, drone’s proximity and low battery alarms loudly screaming.
I had to do something, so I hovered above the bow net and killed the motors. It fell into the net and didn’t damage its propellers/gimbal.
Yes this is what I did. Good thing to know, by the way. Since you don't need it unless you need it, and then when you need it you absolutely don't have the time to look it up.
This command ("Combination Stick Command") is also useful in other situations: the same command can both start and stop the motors. It is exactly what you are going to use when launching and landing in your hand; once you are holding the drone in your hand, and you want to start or stop.
But now: how are you going to do it, when you are holding the drone in one hand, remote in another and do a command with both sticks? Don't you need a third hand for that? The answer is no. You need a lanyard for your remote. That way, you can use one hand for holding the drone and the other for pushing the sticks.
I've flown from boats a handful of times, and it's stressful. When I started, I would ask them to keep the boat as still as possible, hover next to the boat, then slowly move over to me so I could reach it. Grab it from underneath and flip it over to kill the motors.
That's obviously not safe, and it's not hard to find photos in drone groups of people's hands after a failed attempt to do exactly this. It's not pretty.
After that I came across an old indoor practice goal for soccer players - basically a net on a stand. I modified it to lay flat and started using it. I'd get the quad close to the boat, hover a couple of feet over the net, and kill the motors. It worked, but felt like I was risking my quad dropping into the ocean every time I did it.
These days I've gone back to hand-catching, but with a vertical grip that I mounted under the aircraft. I 3D printed a "harness" of sorts that fits around the fuselage and has a Picatinny rail on the bottom (like on a rifle). I picked up a cheap vertical grip designed for a gun that I can quickly mount on the rail - it extends ~6" below and gives me something to grab without being on the same plane as the props. It works well enough, but isn't perfect. It would interfere with the field of view on something with a 360º camera gimbal (like an Inspire), and has a significant impact on flight time and performance even though I did all I could to reduce the weight.
Funny enough I was telling my buddy about the video. And he mentioned he had some youtuber go fishing with him and bragging about some drone he had. He didnt turn off RTH and it went in the drink when the batteries ran down.
You can, for both DJI and Autel (the largest commercial manufacturers). There is similar functionality for the homebrew flight controller firmwares out there (e.g. iNav/ArduPilot).
I'm most familiar with Autel's system, since that's what I fly 99% of the time. It has the option to update your home point, but it also has two modes: RTH to takeoff point, and RTH to the controller.
RTH = "Return To Home"
The problem with all of the above is the "failsafe". If the quad loses connection to the controller, by default it will return to the takeoff point and land. You can set it up to RTH to either the the takeoff point or the last known position of the controller, then land or hover until connection is re-established.
On a boat, you're constantly moving. If you lose connection, the drone immediately loses the current position of the controller, so you have to turn the boat around, manually find the drone, and cross your fingers that you're able to reconnect to it before the batteries die and it descends into a nice, slow, controlled landing in the middle of the ocean.
Identifying alternate safe landing sites before taking off is part of the pre-flight checklist for any flight, but it's especially important when operating over water. Doubly so if you're taking off from a boat. Whenever possible, I make sure to end my flights with enough battery that I can make it back to land if I have to. I've planned to use (inactive) oil platforms in the Gulf of Mexico in an emergency. I have no idea how I would have gotten ahold of the owners and gotten access to it to retrieve my drone if I'd had to use it, but I'd rather work out those details later than lose it or be forced to ditch it in the water.
You can update the home point and have been able to for a long time. And this is zero concern if you're good at manual landing. I've never really ever used their RTH feature the entire time I've been flying, my experience goes back to well before R/C aircraft had any RTH or landing assist.
It's more that you just have to remember to update the home point, and if you don't the drone will suddenly fly off far way out to open water, and if you're not playing it safe you could then run out of battery before you can fly back.
This stuff is annoying, but anyone being responsible and making sure they budget plenty of reserve power ends up being fine.
I've done it on a motorboat and that's freaky enough.
At least with my DJI the "home point" is the trick. As the boat moves you move, but the home point is/was fixed at a GPS location, not a radio location.
So you have to stay on top of updating the home point (which requires menu digging) or the drone eventually freaks out and tries to fly away back to where the boat's route started, which could be a great way to lose the drone.
In the end for me the DJI's level of software nanny plus sketchy china issues have kind of sealed the deal that next time I won't buy a DJI. There are a lot of places that I can legally fly in a responsible manner that DJI's nanny software basically stops or makes extremely inconvenient. On top of places where a low altitude flight would be allowed and they completely block it, their "unlock" process in other warning type areas has been so buggy I've had many times I can't fly even though it's legal because their system isn't working right.
Being able to completely disable RTH (maybe you can?) and or make the home point auto-update to the transmitter location instead of a fixed GPS location would be a lot better.
> Being able to completely disable RTH (maybe you can?)
I believe so.
I've not flown a DJI "seriously" in a couple of years, but my Autels can be set to stop and hover at a given altitude if connection is lost. I'm pretty sure I remember DJI having the same option.
Practice makes perfect- I’ve launched and landed while sitting in a floating touring kayak at sea dozens of times with the Mini 3 Pro. I’ve only knicked myself once so far!
I've done it a few times on sailboats, under sail. I think that's more difficult more of the time because you can't just stand still and land. You pretty much have to grab it mid-air.
I can't see the dang product in the first 5 seconds on the product page. Sure theres a dim shot of it folded if I scroll down (although its covered by popup consent forms), but I want to see the drone clearly without sitting through an artsy video. Something like the second and third picture on https://store.dji.com/dk/product/dji-mavic-3-pro?vid=137691 showing it folded and unfolded are SO much better
is the primary use of drones in the Ukraine war for recon? I know they're also using them to drop grenades and fly into targets. I'm sure every military in the world is watching drone use very carefully. An FPV flying RPG warhead in a dense urban environment sounds like an absolute nightmare to defend against.
It's done in the long range FPV community. Not super common - it's expensive, and you aren't allowed to fly farther than you can see the drone (and safely fly it by sight alone, and monitor the airspace around it) but it's done.
I've not specifically kept up with the long range community, but I want to say I've recently seen videos of test flights out to 100km using an independent radio link (i.e., 5.8GHz analog video + 9xx MHz control link).
That said, power density is going to be your first limiting factor. You're either going to have to seriously tune your airframe for long-range flight (and greatly compromise performance), or you're going to have to move to a fixed-wing design.
Multirotors must maintain thrust (and therefore, expend power) to provide lift; fixed-wing designs provide lift as long as they are moving forward. The flight time of a fixed-wing design that's light enough and has a long enough glide profile is unlimited in practice. You can catch thermals and updrafts to gain speed, altitude, or both.
I have a 1m foam glider that I've flown for well over an hour. The vast majority of that time was with the motor turned completely off and the prop folded flat against the fuselage. I'd soar out over the field where I was flying. There was a creek at the edge of the field and a small hill on the other side of it, creating a "barrel" of air. Flying in low, I'd catch the bottom of that current, ride it up the hill side, then pick up speed as I crested the hill and gently turned back toward the field. That particular location is great on chilly spring and fall mornings, because the temperature gradient between the creek and wind coming over the hillside is so great.
It's been out a while, but is roughly comparable in every way except the camera. The Autel camera has a slightly larger sensor, but only one lens: with the DJI you get a telephoto and medium telephoto lens as well. The DJI will also do Apple ProRes RAW output while the Autel doesn't.
There are other differences, but I'd argue they're relatively minor, and mostly boil down to personal preference.
Autel Evo 2 Pro v3: $2,500
DJI Mavic 3 Pro Cine: $2,999
They're like $200, and will teach you what you want in a drone. That way, when you go to spend real money on one, you'll be able to make an informed decision.
If you can't spare the $200, I'd argue you probably shouldn't be diving deep into this hobby. You certainly shouldn't be spending $3k on this one.
They make them, they just aren't standard equipment. They also reduce flight time and agility, and in come cases can interfere with your field of view.
I've got hundreds of hours flying quadcopters at this point, and the only times I've ever wanted prop guards on a photo drone have been for flying indoors or very near hot air balloons.
No, I have guards, and use them when I feel like it's worthwhile.
FWIW, hot air balloons aren't going to be damaged by the props on the quad that I was flying there - I spoke with the pilots and made sure everyone knew what was up during the pre-flight safety meeting. I used them because I had it set up to hover in place upon loss of connection, and knowing balloons don't have a ton of control over their flight path I didn't want to risk them running into it and knocking it out of the sky. Prop guards would have allowed it to just be pushed out of the way.
Use of the word orcs is incredibly unfortunate as it's clear you're just trying to dehumanize other people. I get why, but I don't think the ends justify the means.
I know the Android app sandboxing should help a lot, but there's enough unknowns & potentially lurking security holes that I really wouldn't want to trust them. They also are 100% the most likely company on the planet to self-update & start truly-epic-level spying on the world if China ever gets a little bit more nervous. It's just not a tenable proposition to trust DJI with this security stance; no one at the company has the power nor right to say no, & as such there's no basis for trust to form.