Hacker News new | past | comments | ask | show | jobs | submit login
Go Where No Drone Has Gone Before (cleorobotics.com)
85 points by Tomte on Nov 27, 2021 | hide | past | favorite | 51 comments



Hobbyists have been building building both coaxial thrust-vectored and planar bicopters for years, so as usual the marketing is overwrought.

I think the success and interesting-ness of this depends mostly on the positioning system - the visualizations in the video make it appear to be inside-out, maybe ToF based on the sensor array? If this is good and the pathfinding autonomy is sound, this could be pretty cool in a pretty small niche space.


"Autonomy: Auto position hold, obstacle detection."

So, no pathfinding.

If it had good pathfinding with no GPS or radio link, it would be an unjammable weapons system.


Exactly my thought as well. I actually find it surprising that we haven't seen more assassination attempts using customer-grade drones, I guess the tech isn't quite suitable yet.


See Slaughterbots [0] for a dramatization of this scenario.

[0]: https://youtu.be/O-2tpwW0kmU


There has been a few if you monitor the news, mostly unsuccessful (e.g. Iraqi PM a couple of weeks ago)


Here's an idea I had recently - are you aware if anyone's tried this? An aircraft with coax rotors. Off-axis torque is achieved by changing the angle of the blades independently around the rotor. If thrust vector is in Z direction, and you want to rotate in XZ plane, you increase AoA of rotors in +X direction and decrease AoA in -X direction, causing differential thrust which rotates the aircraft from pointing in Z to pointing more towards -X. Control would probably be via 4 or more cams around the shaft.


I think you're describing a collective pitch coaxial helicopter - these exist in full size as well.

But by your description, with more fine-grained control than a swash plate, maybe?


> collective pitch coaxial helicopter

Yep, that's it! I see some drone videos with exactly what I was imagining. Russian KA-52 looks like full size version.

I didn't know about swash plates but that's basically what I had in mind.



Super cool. I am just getting into drones, and I am surprised at the flight time limitations of most commercial drones. This unit for example, to scan a mine network, would take constantly repositioning an operator to charge the unit as you make your way through the site, as the reach required will get longer and longer as you scan more of the site.

What in others opinions is the fix for this? Is it a technology that doesn't exist or is it a technology we aren't allowed to use (tiny nuclear?)

Of course it's always a design compromise, there are huge drones with 40min flight times, fast drones with 5min flight times, and much inbetween. But I guess I see the practical application of drones challenged by the flight time limitations.


It seems like it's just a fundamental rocket equation type limitation, battery tech has a very low mass vs energy density, and it's paired with an aircraft that will always be much lower efficiency than fixed wing.

I'm sure much longer flight times could be achieved using combustion engines, but obviously that wouldn't be a great idea inside a mine or in sensitive areas.

One workaround could be a sort of blimp/dirigible, but I don't think it could be scaled down the same size as a quadcopter.


The problem is that you have to have heavier batteries for longer flight time, and then you need more power to lift that extra weight, so more power needed, so more battery needed. It's a vicious cycle. I'm sure there's some rule or law about it.

And unlike normal aircraft fuel, battery-powered drones don't get lighter during the flight. Batteries weight the same charged or not.


In my opinion the fix is to stop using rotary wing designs, they're simply inefficient.

Instead, use fixed-wing designs. There's a reason planes are so much more common than helicopters.

Another alternative might be very small airships, especially indoors.


Planes need forward velocity to generate lift using fixed wings. No vertical takeoff/landing/hover with fixed wings unless you do some very fancy thrust vectoring.


Teamwork; you have another drone carrying batteries or whole drones in and out.


Interesting that there is a new entrant in this market, there was already https://www.flyability.com/ for indoor inspection.

A lot of other comment seems to be mentionning that it's a niche, but I think it might be the biggest market after all. The retail market for drone is already saturated and the competition is mainly on prices. In that case, firefighter equipment, the market is global, huge (every small town has firefighters) and firefighters are much more deep pocketeted than end user, are ready to pay much more for high quality equipments.

Now the challenge is to convince every firefighters that they need one.


Looks like it can fly for 12 minutes on a single charge and takes 40 minutes to recharge. The video is super neat, but I'm curious how loud it is.


Looks like it's for industrial processes so it probably doesn't matter how loud it is. Like photographing the inside of a storage tank for defects, where you won't get a reliable GPS signal (or likely any other RF).


I wanted to hear how loud it is, but they added background music instead.



As well as noise (I'd presume it sounds somewhat like a leaf blower, particularly given the video has no actual sound of the drone), I'm wondering what sort of interaction it would have with floor/walls/ceiling in tight spaces (this would be true of any drone). And if there's dust (as hard-to-get-to places tend to accumulate), what impact that would have on, among other things, the camera/lidar.


Having done some drone flying in tight quarters, I'd say the number one issue is situational awareness. That this is resistant to bumps is great, but you can achieve that with prop guards on a normal drone. What you need is a dedicated camera pointing straight up and one straight down, live at all times, in addition to your normal gimbal.


$10K!


Ouch. This outright kills my idea to use it for my cat's entertainment.


Mavic mini is on a pretty good Black Friday deal at Amazon right now if that’s what you’re after. I just ordered one.


I already have mini-2. Did some nice filming last summer. But it is no fun to use indoors at all.


Get a tiny whoop


My drone is 75cm cube and go through pipe with a tiny camera and GPS denied environment.


75mm?


Oh my bad 75cm is almost a meter. Yes it's 75mm toy drone.


Really solid. I am excited about this one. Good design with clear problem to solve.


Range: 1km

Flight time: 12min

How is that "good design"??


That’s a pretty interesting looking drone, but from the title I was hoping (but also knowing it wasn’t going to be) for more drone through volcano footage or something equally radical.



In thought they were referring to the Mars drone. But Ok. Tell me when it's flying in another world.


I guess the main use is meant to be for inspecting industrial machinery? I feel like they should have had a demo of the drone bumping into an epo button (or a button that operates a machine) to show it doesn’t get pressed. I suspect the drone would be too light but it would be a shame to have your kit turned off by a clumsy drone with a silly name


I think momentum could get it pressed, but this thing is packed with area sensing so it would be surprising for it to hit anything when being operate autonomously and sensibly.


If it won’t bump into things, why do they have a video of it bumping into things with a subtitle saying it won’t do damage?


Human operators I suppose. I'm not saying it's not a great feature to be able to bump into things, it's probably the entire design goal of enclosed coaxial design even. But the intention isn't to bump it's way down a corridor, it's still got object avoidance.


It looks a lot like Weebo from Flubber. I wonder if they based it off of that design.


Interestingly their speed of 4m/s is software limited rather hardware!


How is the control done? Are there servos controlling the output ducts?


Could be really interesting for police or search-n-rescue operations.


Also for search and arrest, mobile surveillance and drones in every damn corner of our cities once these get mass produced. They don't need to be flown by a pilot, there are already companies that work on autonomous drone software. And then the data can be analyzed in an automated manner as well.

Think about the possibilities: No more searching for fugitives for months, the drones simply go around our cities and check every single person. 5G, wifi signals, infrared cameras and other sensors give them the coordinates of all humans. The state will be all-pervasive. Laws and standards are for this are already created as we speak.


12 minute flight time is terrible. Like the fan design though.


Very cool, but can someone’s hair or clothing get sucked into it?


>> No exposed propellers and a frame build from composite materials means the Dronut® is tough and ready for duty.

"Duty" - presumably that means it's meant for use by imilitary or police forces.


> Flight time 12 mins

> Wireless link Wifi

> Range (environment dependent) 1 km

LoL... I'll pass. When a DJI Mini2 can give a 10KM range, why would one go for this?


In practice, the DJI Mavic’s I’ve flown never make it that far can with LOS


area 51, here I come


Range: 1km

That's not taking you too far into Area 51, I'm afraid.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: