Lily – Drone camera

1397 points by duck ↗ HN

494 comments

[ 3.4 ms ] story [ 289 ms ] thread
(comment deleted)
That's really, really nifty. Was wondering when a "point and shoot" drone would hit the market.

I really want one but I've grown wary of pre-ordering products that haven't been manufactured yet...

Thank you Nadya! We are shipping in February 2016 and will be sending out frequent updates. Stay tuned at www.lily.camera.
That is one of the first videos I've seen in a long time where I've thought "wow we can do that now?"
(comment deleted)
Thank you! This is just a glimpse of what one can do with Lily. Check out our website www.lily.camera for more information.
The submission is literally a link to the website.
This is some of the best marketing i've seen in a long while.

Just in case it's not obvious: I'm not complaining or being sarcastic, this is honest praise.

Brilliant video. Felt the need to watch it twice through in quick succession.
Wow! Preordered, but used my AMEX in case it turns out to be VaporWare.
(comment deleted)
Out of curiousity, what does your AMEX card offer if this is VarporWare?
Amex allows you to dispute charges very easily. Their dispute resolution process is slanted in favor of the cardholder rather than the merchant.
Is this a reason that many merchants don't accept AmEx?
Partially, yes. But Amex also charges higher transaction fees, which come out of the merchant's profits. Between the two, and the fact that only ~20% of people have an Amex, many merchants decide Amex isn't worth the trouble.
Re: shortstuffsushi

Generally, merchants don't take AMEX because it's expensive to do so. The fee rates are usually higher (~3.5% vs ~2.5% and there is a monthly access fee as well.

Additionally, Amex provides both the credit as well as credit card processing for their customers instead of splitting the business with thousands of banks. So when a merchant exceeds whatever shadiness threshold is set by Amex they are banned from accepting all Amex cards vs. cards issued by $yourbank. So there is incentive for the merchant provide good customer service.
Oh, well AMEX in general is the most liberal with consumer-protection. If it ends up being fake, or the end product is nothing like the description, or whatever, AMEX just credits you with no questions asked. They also do lots of cool stuff like automatically doubling warranties and such.
This is cool. I've been dreaming about something similar. In my mind though it would be smaller and could hitch a ride on your shoulder.
I imagine that one problem is that if it's too small, it will be unstable in wind... So since they need a certain amount of weight for stability anyway, might as well devote it to big batteries (rather than being small and dense).
You're right. I wait for the day when we can make the batteries smaller / motors more powerful.
This is great. Now, we're just waiting for better batteries: stretching out 20 minutes of flight time across a day or even morning of skiing, kayaking, etc. is going to be difficult. I guess you can carry batteries with you.
Especially at the rate that batteries degrade. It'll be no time before you're getting 3-4 minutes per charge!
Battery is not removable.
That's horrible and possibly a deciding factor. I guess it had to be done in order to make it waterproof.
I concur, you really need the possibility of multiple battery loads in order to get full value out of this stuff.

Can you do non-conductive power transmission at reasonable weights and efficiencies? I mean, my toothbrush does it, but I can't speak to the charging rate.

Wireless power transmission can only be implemented using alternating current through induction. Very hard to implement this in any meaningful way for battery powered digital devices.
These are cool, but...

1) 20 minutes is great if your sole purpose is getting video, but not if your sole purpose is surfing (let's say). A surf session might be 2 hours. Taking a break to put your dead flying camera away in the middle of it isn't ideal.

Still there could be work around for this at some point.

2) Air space. The real problem. If you're surfing (again) or skiing you'll annoy everyone with the sound of your drone.

Still for the more adventurous who get to surf or ski alone, this does look pretty awesome, but will never be as ubiquitous as the GoPro.

This drone does not seem at all to be optimized for surf. It even says in the FAQ to be cautious when putting it in salted water.

But yes, 20 minutes is really short, it's its major default. I suppose it was a necessary sacrifice to make this kind of product at this time.

Charging base where it could land and somehow automatically change batteries (or switch with another drone?) would solve some of this. Hopefully battery tech will pick up a lot in the next 10-20 years!
It's waterproof, but still the lens will get wet. How did they solve that?

Also, does it have built-in collision avoidance?

Probably have a coating of hydrophobic treatment so water slides off.
Solve what? The lens gets wet just like any other waterproof camera. At least this one will blow dry itself off in short order.
In the video they looked apprehensive for a moment when lily landed in the river next to the kayak. Or maybe that was just me feeling apprehensive for them.
Serious question. Is this not a thing being sold already? The design is really nice and the video is really well done... the waterproof seems maybe a new thing.

I have seen multiple promo videos for various drones offering same "follow me" feature.

Here's one I found in few seconds https://store.3drobotics.com/products/iris

The 3d mapping on this one is amazing.

I wonder how this handles obstacles. The trails we ride are not wide open ski slopes, or rivers, often times they are tight/wooded trails. Would like to see how it handles "close following".

Lastly, the part in the video where the guy just tosses it into the river was definitely mind-blowing.

This is the big reason people don't buy these types of drones (this one certainly isn't the first of it's kind).

It doesn't have obstacle avoidance or it will make a massive deal of it.

Intel last keynote demoed very capable drones. They're showing it off, it won't be long until it hits the market. They named the components RealSense, a miniature kinect I guess. Nice demo at 4:55:

https://www.youtube.com/watch?v=Us0BqJvsF9k&t=295

RealSense uses projected IR and thus doesn't work outdoors.

(edit: because short-range and little ambient IR are required)

Thanks for the tip. Are there any portable technology that would work outdoors ?
Vision works for animals, and a lot of people are working very hard to do it for computers. Google "UAV visual SLAM" for examples.

LIDAR works already but is heavy, power hungry and expensive. Not good for these small vehicles.

Yeah, I pictures a Google Drone laughing.
stereo vision, structure from motion, animals use both
Except the end of that video shows it working in a forest... https://youtu.be/Us0BqJvsF9k?t=6m14s
OK, that's cool. I stand corrected. Clearly it can work outdoors in some circumstances. IR texture projectors get washed out by ambient IR from the sun. Thus the choice of a shady forest for the very nice demo. People have also used Kinect outdoors in the evening, etc.

But in the general case you can't rely on this approach because sun.

> because sun.

That's a fair point. So RealSense will probably work at night, in a shady forest (like the demo) or maybe the shady side of a mountain, but as soon as the sun is involved IR becomes unusable. I don't suppose you could switch the other side of the visible spectrum and do UV cameras, could you? Having your data reduced to one dimension per pixel really makes the AI easier.

There's a lot of UV in sunlight too. Hence sunscreen. Many birds can see it. Some birds have patterns on them that are only visible in UV. Fun facts.
I probably have no clue what I am talking about, but wouldn't it be sorta like steering forces?

Seek A Avoid B

In your case, a ton of Bs in terms of leaves, branches and etc.

I guess it would need a wide scan to detect if there is a path either vertically or horizontally.

Well, (A) has a tracking device. All the (B)s don't. If they did, then yeah, this would work.

Right?

Yeah but I am thinking it has some sort of an scanner, sorta like self-parking cars do :).

So once it Avoids (Steers away) B it can continue to Seek (Steers toward) A with the tracking device.

Obviously depends on how well the scanner to pickup obstacles work. You would need some sort of a heat map to detect leaves or branches.

It is a hard robotics problem. Saying that "it would need a wide scan" is a vast underestimation. Try looking up some papers on obstacle avoidance. Also, consider the costs of sticking the additional sensors on the robot, processing that data in realtime, and still powering the whole thing.
I am guessing the biggest problem is the noise. Yeah, it's easy to figure out if there is a giant wall standing in front of you, but what about snow that constantly blocks its lens? Some leaves may be attached to tree so it might be required to circumvent, but some might not even be worth circumventing if it's just a piece of leaf falling.
maybe someone can pull it off with ultra sound? Those can be pretty cheap.
Probably won't have much resolution for leaves. I've used ultrasonic transducers for a college class project almost 15 years ago. Ultrasound worked great for hard surfaces like walls, but softer materials like clothing reduced its range of detection. Snow covered trees, I think will have a similar signature that would be hard to detect except at close range. With a UAV, it could be done, but it'd have to be moving pretty slow I think. You'll probably have better luck with some sort of optical system.
I guess the drone wouldn't be able to launch from the water if it had landed upside-down?

(As a concept, this is amazing - excited to see how well it works in practice.)

Waterproof will be amazing safety for kitesurfing and surfing in general, quadcopters are super cool, but you don't want to drop one in salty water. There are already techie stuff for kiters, we use the Woo device and gopros a lot.

Most areas have a club to keep members under control (irresponsible kiting can be dangerous to people not in the sea, unlike surfing) ... our club is keen on getting Woos, GoPros, webcams, internet enabled wind-o-meters etc... Hope they decide to get one of these!!

I can't imagine that propellers play nicely with kites...
if you could but a beacon of the leading edge of the kite and setup a NO GO zone between the remote and the kite's beacon that would solve that issue.
Sounds good in theory, but in practice kites can move really fast when manoeuvring or doing tricks. No guarantee that the drone would always avoid you successfully. You'd have to stay aware of where it is and where it's going, which would probably interfere with your kiting significantly.
Waterpoof also means we can start seeing drone footage from within rain storms.
I would think the wind in a storm would pose a challenge, but I'd love to see that.
The biggest showstopper for me on this is the hugely disappointing top speed (40 km/h). I'm an alpine skier (ex-racer) and the only times I go that slow is on transport stages. This thing literally wouldn't be able to keep up with Usain Bolt running the 100m. It needs at least twice the top speed.
Usain Bolt does around 38 km/h
Yeah, but I bet he accelerates faster.

Edit: he accelerates from standstill at 9.5 m/s^2. I couldn't find data on what drones do.

Drones usually accelerate quite rapidly (electric motors and all).
Well electric motors can accelerate rapidly, but if they're pulling something heavy, all you'll get is rapidly spinning electric motors and a slowly rising object.
Still, that's zero to 95% of top speed in one second (or 5 meters).
Not for 20 minutes.
That's actually pretty speedy for a quad, especially to remain stable. If you want to race at 100MPH and have gyro-stabilized video you should get a helicopter and a camera crew.

I for one would prefer the aerial robot following me to not be traveling at 60MPH. That's asking for a serious injury.

Not to mention that the speed (+-) is variable by the wind speed.
Well, I would imagine a lot of your speed is downward right? Then it's a matter of a controlled decent really. I'd guess that would increase the top speed of the device as it's really just falling out of the sky and steering.
It's electronically limited to 40 km/h, so I don't think so.
Maybe you can send it ahead of you and instead of following you it just shoots as you come towards it?
Wonder if you can buy a few and tell them to follow each other.
The FAQ says they're working on having multiple cameras follow one person
This is the true hacker mentality right here
You could probably just duct tape the tracking device to another drone. Looping them together like ants would be cool! :D.
Tape the last drone's tracker to the first drone, and launch them all at once. I'd pay to see that video :-)
How much would you be willing to pay? For about 5 grand I could make this a reality some time around February 2016! Interested?
(comment deleted)
Yeah. I kiteboard and I would love to make some images using Lilly. But I'm pretty sure eventually Lilly may crash into the lines or kite.
Same. Also curious about the max wind speed it can handle
Ugh thinking of that gave me shudders. Kiting is dangerous enough without these things flying around, I would definitely give someone a piece of my mind if they started flying a drone in a busy spot.
I wonder how well it will work for surfing. Probably great for small crumblers and open faces; probably not so great at the Right: https://www.youtube.com/watch?v=_N_-vxfz4uw . Can you imagine how sick it would be to get this thing to follow in a tube at Teahupoo: https://www.youtube.com/watch?feature=player_detailpage&v=7w... !!
do you surf? Have you ever experienced the push of air that develops from a closing / collapsing tube?
yes, yes [i assume you mean either getting spit out, or closed out on]. drones are starting to make inroads here too, but it is still mainly remote stuff, taken from afar and manually controlled by someone on the beach. something like this at jbay would be amazing.
I see it the other way around. This could help humans handle obstacles, at least if it could launch itself automatically and when it's a bit smaller so that it's really easy to carry around.

African American males could carry it. Then, when police was nearby it could launch itself. Maybe it would be triggered by the sound of sirens or if there is a website that tracks police cars, maybe it could launch every time they are near. Or maybe an apple watch could detect fear from differences in pulse and then launch it. Maybe some other trigger

A future, self launching, slightly smaller version of this could be a great protector of civil rights.

This is a colossally stupid idea. How is this better than someone having a cellphone? Are you always holding it outstretched on your hand, anticipating a cop to drive by any second? Or can it crawl out of your pocket to defend your civil rights?
Or perhaps mounted on top of an ACLU baseball hat?
This is not a bad idea.

Women could carry it if they run at night and if they don't make it home it could fly to a police station.

Parents could send them out with their kids.

(comment deleted)
That would enable the truest form of helicopter parenting (https://en.wikipedia.org/wiki/Helicopter_parent)
I heard the phenomenal term "curling parenting" the other day, where the kid is the stone and the parents are the two people frantically working to reduce the friction in front of it.
(comment deleted)
(comment deleted)
Police do not react well to being filmed. What will likely happen is that your drone will be confiscated as "evidence" or destroyed. Its not as if cell phones do not exist to fill the same role here.
The ACLU makes an app that streams the video recording to an Internet server in real-time at the best possible quality given your data speed. It will also upload the full video when possible without additional action.
Im sure you can do that with a phone. Im not sure what the drone adds to the equation.
> I wonder how this handles obstacles.

According to the creators v1 doesn't handle obstacles at all due to time and cost. I'm sure a later version can be made to do so; I feel like it should be far easier to do in a 3D space versus a flatted space like ground vehicles.

Relatedly, I wonder what would happen if you turn it on and then walk inside a building...
> rivers

I'm a rower so fortunately I won't have the obstacles issue. In fact, these could be great for rowing coaches who can put the controller in the boat then get close up to the rower/crew without needing oversized launches that wash everyone else down.

20 mins might be a problem, but I don't see it as insurmountable.

I'm very impressed by the capabilities. I've seen 'follow me' drones before, but I don't know if I've seen any that are waterproof or as compact as this. And the smiling face on the front is a nice touch.

It's still hard to part with my money (even at 50% off) when the ship date is next year though. I've been burned by various kickstarter projects that get delayed ad infinitum.

I've been burned by various kickstarter projects that get delayed ad infinitum.

Me too. I've decided that there are no new things that I need so much that I can't wait until the thing is actually shipping. In many cases I've been quite happy waiting until a v2 release when the early bugs and quirks are worked out.

This is exactly why I've grown wary of pre-orders.

Even with huge savings for being "an early backer" - I've found I get more for my money (and sometimes anything for my money) if I wait until an improved "v2" ships a few months after initial release.

From the FAQ: "What is your policy on refunds? If you decide to cancel your pre-order before Lily ships, we will provide a full refund to you."
Assuming of course that they don't run out of funds before they ship.

This isn't a slight on this project specifically but there is a risk that any project which isn't yet in production won't get that far...

And if you believe that, read up on the drama with https://www.kickstarter.com/projects/pirate3d/the-buccaneer-...

TLDR; during development of the printer, a number of things have changed (including a noticeable reduction in print volume). Refunds were offered August of last year, and the vast majority of refund requests are still pending (and currently half a month behind their own deadline for the latest update.

Not that I have any knowledge on whether the people behind Lily have better or worse communication and management skills, but just because they're happy to offer refunds doesn't mean they'll offer them promptly.

Kickstarter is different than a direct purchase. Kickstarter maintains they aren't a store so when things don't happen to plan you're kind of on your own. With a direct purchase you can do a chargeback easily and get your money back one way or the other.
> With a direct purchase you can do a chargeback easily

Not after 90 days you can't.

I'm curious how it homes in on the tracker.
Apparently based on the hand device.
i think the question is what tech does it use do accomplish this
Yes this is what I was wondering. Some brief searching on similar devices suggests it's a combination of precise sensors and algorithms to tie everything together.
Without a human operator assessing risk, what's to stop the drone from accidentally flying into someone and cutting them up with its blades?
There is typically not enough force in the rotation of the blades (they are designed to be very light) to do more than hurt a little bit.
I can't speak for the Lily blades, but I have a couple of drones a little larger and the blades are harmless - they stop as soon as they detect obstruction and they are flexible/not sharp. It won't cut anyone.
I just got the DJI Phantom 3 and got hit by the blades after a collision. Sliced up my arms -- didn't require stitches, but 10 days later i still have evidence that the blades aren't totally harmless.
There are plenty of gory images on forums of people who foolishly think this is the case and are proven otherwise!
I have NEVER seen a controller that would "stop as soon as they detect obstruction" like you say. In reality the flight controller is likely to do the opposite. Once it detects that the current speed of the blade is not enough to provide the desired lift, it will INCREASE the speed.

Telling people the blades are harmless is very irresponsible.

The electronic speed controllers we use in RC have a loop time of a few milliseconds and are fully programmable. Some of them are capable of sensing the current going into the motor. You can also monitor the voltage instead of/in addition to the current. It's easy to cut the throttle when there is an unexpected power surge.

I personally haven't seen this technology in DIY multirotor parts, but I know that the AR.Drone has what they call "cut-out detection". They use it to prevent damage to the craft rather than humans, but the technology could be tuned for this as well.

The most popular low end quadcopter, the Syma X5C, has this. The blades are flexible and stop almost instantly when obstructed. This is great for preventing breakage when you hit a tree. I've stuck my finger in there intentionally without even mild pain.

If a $60 device has it I bet the multi-hundred and thousand dollar devices will not have trouble.

I see multiple comments saying that the blades are harmless, but a quick search through google photos seems to imply otherwise to me. Not to mention the force of this thing dropping on your head after it clips a nearby branch, or has a system failure.

I suppose the answer is that the owner needs to be responsible with its usage. The example footage seemed fairly reasonable use-cases (no bystanders).

'harmless' drone blades are the exception, if they even exist. Note that drones designed to operate indoors have some form of protection for the blades (Parrot drone, for instance).

In addition, 'cheap' blades have sharp edges due to the manufacturing process. Some drone owners recommend sanding them to get rid of those, to minimize injuries.

I could think of a pretty much harmless blade design that MIGHT work (soft blades), but they'd be woefully inefficient AND they would only work for those super-ultra-light micro/nano drones.

Anything else is not harmless. Even a tiny quadcopter can easily cut your skin, because the plastic is hard and sharp. Think of it as one of those plastic picnic knives, except not that sharp --- BUT rotating very, very fast.

Yeah, those blades are NOT harmless. They will cut you up on the thinner skin bit and do some serious bruising on the thicker skin. Also, anything around the head is asking for losing any eye or something else.

I've been building and flying quads for years now and Anything over the tiny indoor ones commands a LOT of respect from me due to experience.

The idea of these things autonomously following you is cool but a HUGE liability.

Unless they have some major obstacle avoidance on board they you basically could never safely use this around other people and honestly you shouldn't even use it around yourself.

Video made me want to have a more active life. I thought "cool!" but then I imagined my video output would be hours of me typing, browsing HN & Reddit.
Somebody should do a spoof video of their marketing video that shows exactly that.
(comment deleted)
Somebody should make Office Space 2. http://www.imdb.com/title/tt0151804/
HBO's Silicon Valley series is probably the closest thing we're going to get to Office Space 2. Which is good, because the series is genius
It's from the same director. I like the series, the second season in currently on TV.
Sad realizations are the best inspiration.

But really, working in internet tech, there's no excuse not to get out more often. Internet connections are everywhere and portable computing is absolutely viable.

google NOLS, outward bound, or "mountaineering course"
Yeah. "It could follow me whilst I mow the lawn!"
One feature (or use case) that I would like a follow-me drone for is path-lighting.

That is, let's say I'm going on a walk in the dark and I don't want to use a flashlight or wear a headlamp - I've thought of a drone flying 8-10 feet up, a few feet ahead of me, shining a light to light my way.

Since the Lily people seem to be here answering HN comments, I wonder, how far off is your product from doing that ?

Duck tape and led torch (flashlight).
You'd rather use an expensive, noisy drone that works for twenty minutes than a $15 headlamp that runs for hours?
Sorry, what? How on earth would this be better than a headlamp, apart from the "look at my drone, my drone is amazing" part?
Well right now it's totally impractical, but if it floated silently, it would be pretty amazing. It could even light the path that you need to follow to help you navigate, and it wouldn't cause you to blind anyone you turned to talk to.
"We choose not to do these things because they are practical, but because they are cool!"
> "look at my drone, my drone is amazing"

This is actually a fairly compelling reason to do things.

True dat. But then the lamp part is just extra weight that also drains electricity from the batteries.
...while keeping the user perpetually in the spotlight.
It can also carry my backpack. Like my own army of air sherpas.
I am inclined to use a headlamp for this almost always.

However, if you are walking with one or two other people, their light depends on the direction of your head. And also on your relatively close proximity to them.

I dunno. It has occurred to me several times.

A 'headlamp' doesn't seem that useful. A 'search spotlight', like the police helicopters, could be more interesting.
Wow. This is purely amazing! Definitely ordering one.
Please leave the scrolling behavior alone already!
Nice. We were doing sailing last week and was quite hard to figure the whole activity with a GoPro, because I was focusing only on what I was doing, but my fellows trimming the sails were not that much in the video.

So I was thinking of a drone to film us from the side or above, but our skipper thought that no drone could handle the 15-20 knots with 30knts gusts winds we were sailing in.

There are definitely drones that can fly way faster than 30 knots, but they are not waterproof, and the ones I've seen are rather dumb, they don't follow you.

Here's a drone that can fly at 100 mph (87 knots):

https://www.youtube.com/watch?v=8p5uDf9i_Yc

That's oddly scary to watch for some reason. I'm not sure why?

Maybe due to the fact drones are primarily used for war at the moment...

Primarily? I'd highly doubt that. Small multirotors and model aircraft like these are far more ubiquitous than the machines weighing upwards of 2,000 kgs which are used by the military.
Regardless of form-factor the military is the primary player in the drones industry. The vast majority of operations and spending are classified.

I was thinking more about the future applications of high-speed small drones when they become ubiquitous. Even if it is simply reconnaissance.

I don't think multirotors are used for war, except perhaps for reconnaissance.
Drone is way too generic of a term. Do you know of any military application of a multicopter?
Well, is not about the speed of the _drone_ but being able to fly in a 20knts wind, with 30, maybe more, gusts. Which means in order to keep course it needs to know the wind speed and compensate. It's been done ( for auto-boats ) but I doubt this drone has this.

Last time on Hintertux at 3500m we had also strong winds while snowboarding. I'd give it a test nonetheless!

Right, but all wind is to a drone is a shifting reference frame. In order to stay still in a 20 knit wind, the drone would need to be able to have a top speed of, at least, 20 knots.
It actually is about the drone being able to fly faster than the wind. If it can do it, it can go in any direction it wants (though slower against the wind).
Damn, that's impressive. I want to know the top speed and what kind of wind it can handle.
Just idle speculation, what happens when 30 follow-me drones are on the slopes at the same time? Will they all be able to avoid each other?
The minute drones like these start shipping they'll be banned from all ski resort faster than you can say pow.

Also, with a top speed of 40 km/h this isn't able to keep up with anyone who's past the rookie stage skiing/snowboarding. So their marketing is very misleading, either by intention or because these guys don't know shit about skiing.

Anyone going faster than about 40km/h isn't really doing anything worth filming anyway, other than ski racers.

All the interesting stuff for filming on the slopes happens at relatively low speeds - park, moguls, jumps, butters, etc.

If you're going any slower than 40 km/h the only thing you'll be doing in a terrain park is falling on your face. This [1] cable-driven terrain park "for children and adults" pulls you at 40 km/h. If you're looking at landing anything bigger than a 360 you'll need to be going closer to 60 km/h. Ski racers are not happy until they're going (a lot) faster than 90.

[1] http://www.al.no/en/html/aktiviteter/vinter/ski_snowboard_ca...

I'm a level 3 snowboard instructor in Canada, and I would say I infrequently go over 40km/h. It's only when I stand still on a groomer and "go fast", or for the run-in for the L/XL jumps.
Have you actually measured your speed though? Because the difference in speed sensation between a snowboard and a car (or bicycle) is just unbelievable if you haven't measured it.

The pull rope I linked above is for kids with small/medium jumps, and that's 40 km/h.

Yep, sometimes we race with GPS to see how fast we can go.

80 km/h on a snowboard is fast.

That's interesting. Way back when I was a ski racing kid (about 14) we once did a speed test where we closed off 400m of one of the children's slopes and put a photocell at the bottom, and we would hit 90+ km/h on that thing. Mind you, this was in racing gear with racing skis, but that was a flat-ass slope. The fastest we ever clocked in a race was around 120, when I was 18, in the worldcup downhill course in Åre.
Yep, having two planks and two edges on the ground is a huge advantage.
Like GoPro, just 10 times better?