Skydio's Forthcoming Consumer Drones Can Sense and Avoid Obstacles (technologyreview.com)

moon_unit2 writes: DJI's new Phantom 4 drone may be able to stop if there's an obstacles directly in front of it, but MIT Technology Review has a story about a much more sophisticated self-flying drone, from a startup called Skydio (basically using high-speed visual SLAM, which is no mean feat in such a tiny package). The company's prototype uses several video cameras to navigate around obstacles at high speeds through busy airspace. The technology could make consumer drones much harder to crash, and it could let drones do more complex surveillance tasks. Skydio, founded last year, has so far raised $25 million in funding in a round led by Andreessen Horowitz and Accel Partners.

Comments

So what's the technology

[buck-yar]

Maybe I missed it in the article, but are they using optical flow, LIDAR, IR-lock, sonar or something new?

Re:So what's the technology

[JoeMerchant]

Likely IR, here's another: http://www.onagofly.com/

Re:So what's the technology

[Anonymous Coward]

It says right there in the summary that they are using "visual SLAM" and "several video cameras"; if they were using something other than visible light video cameras, I don't think the summary would have been worded that way (assuming the summary is accurate of course), so no IR, LIDAR or SONAR.

Re:So what's the technology

[JoeMerchant]

That must be awesome when a cloud of dust or blowing leaves enters the visual field...

Re:So what's the technology

[I4ko]

SLAM is a process, not a technology. It still requires sensors. If they are using cameras, they must be doing some funky 3d vision thing.

Mine is bigger than yours!

[Anonymous Coward]

Aaannnnddd, the MQ-9 can blow shit up! What is this, a VC slashvertisement? DJI has made a REAL product and sold a ton of them. Yes, this cool hip company funded by douches in Silicon Valley can make a sample size of one prototype, that is awesome, I get it. Tell me when they can make hundreds of thousands of them at less than $500.

What about buckshot?

[Anonymous Coward]

How is it at dodging buckshot?

Re:What about buckshot?

[Voogru]

Some people may think they can outsmart me, maybe, maybe.

I've yet to meet someone that can outsmart BULLET.

Re:What about buckshot?

[U2xhc2hkb3QgU3Vja3M]

https://www.youtube.com/watch?...

Ceiling detector?

[Anonymous Coward]

My biggest challenge operating a micro-drone is that enough throttle to lift the thing off the ground is enough to slam it into the ceiling. It would be nice if they could automatically throttle-back when near the ceiling.

Now that people are working out how to have these things fly themselves, it's time someone designed swarms of drones to knock dust from tables and shelves to the floor, which would then be picked up by a Neato or Roomba.

Finally a drone that avoid obstacles

[Anonymous Coward]

Finally a drone that can avoid my very big dick so I don't get hurt when flying one.

Normal

[nospam007]

Without that feature these 'drones' are just remote controlled copters.

Re:Normal

[neilo_1701D]

Without that feature these 'drones' are just remote controlled copters.

And?

Without the plethora of onboard computers and sensors, the Google self-driving car is just another crappy Prius.

I don't see the point you're trying to make.

cheap cameras?

[Anonymous Coward]

Looks like basic cameras in some visual stereo configuration--depth map generation makes sense for SLAM. RGB-D would have made more sense as that's a lot of cameras. That's means no flying in the dark.. Then again, I see some terrain following and visual tracking. As for collision avoidance, the example could have been just the former and not any avoidance path planning.

The Ascending Technologies (now Intel) real sense solution looks more practical.

Having collision avoidance isn't the holy grail though a good thing--aerodynamics can still bring the airccraft down (like getting too close to a wall)--and you can't change those Physics.

Zano reborn

[bored_lurker]

This sounds like the rebirth of Zano, the biggest failure in Kickstarter history. For those who don't know Zano raised $3.5M and delivered nothing to its backers. It was suppose to be a self navigating drone that used camera to navigate around obstacles (sound familiar?). They had slick videos of what it was suppose to do, and in the end they delivered - nothing. It was such a disaster that Kickstarter actually hired a journalist to chronicle what happened to the project (https://medium.com/kickstarter/how-zano-raised-millions-on-kickstarter-and-left-backers-with-nearly-nothing-85c0abe4a6cb#.imzfeeu3j). Seriously, why is this different?

Re:Zano reborn

[Anonymous Coward]

Why should it be the same?
Other than the fact that it's a superficially similar product - is there _anything_ about this that's similar?

Re:Zano reborn

[Anonymous Coward]

Further to that - the very first paragraph of the article shows an immediate difference:

"On a recent bike ride through the woods near Menlo Park in California, Adam Bry, CEO of a company called Skydio, and his coworkers were joined by a small, nimble drone. As they rode along a dirt track, the drone followed close by, weaving expertly around tree trunks and branches in a series of deft maneuvers."

The reporter literally saw it actually in action, in the wild. That's more than anyone (even employees) ever saw of Zano.
In addition, this is a genuine research project. It's not an attempt at making a cheap consumer toy - from the picture that thing has to way a few KG.

It's also not a Kickstarter project, and has $25m of private funding. There really, truly is nothing even remotely similar between the two.

I do not think that

[thoughtlover]

means what you think it means.

"...using high-speed visual SLAM, which is no mean feat in such a tiny package..."