5 ms·

Can Airware be the operating system to connect the world's drones?

39 points by katm 11y ago

amelius 11y ago
It sounds like vaporware :)
- sjtrny 11y ago
  It is just a pr piece.
  - evilolive 11y ago
    It is definitely PR. Airware has invested a lot more into doing its own hardware, its own possibly certifiable low level code than doing exciting things (sense & avoid etc). Heck, no amount of money will let you write a single line of code on their OS for drones currently. But is the PR going to work ? It seems easy enough to deter the "serious guys" doing some "serious commercial inspection" from using 3dr. The same way people use cisco equipment on their network because they're told that's what you do if you want peace of mind.
andymoe 11y ago
I don't think the operating system to connect the worlds drones is going to be closed source. 3DR and their open source base is going to give Airware a run for their money as the de facto standard "operating system for drones" in the consumer and more and more in the commercial space. In the meantime people like Drone Deploy and Skycatch will fill the gap when it comes to data collection and processing which is where the real business is long term. Drones themselves will continue to become a commodity. Also, it's worth remembering that a Chinese company, DJI, has about 80Pct market share in the US for consumer and prosumer drones. Lots of competition for everyone.
- freeqaz 11y ago
  3DR's software doesn't compete in the commercial space. It's a hobbyist platform and it will be much harder to certify by the FAA or other agencies. That is a huge portion of being successful in the space. Especially when you're doing things that have more risk, like photographing electrical infrastructure in a densely populated region. The FAA is going to be the agency that determines if you can do that, and the companies that has that advantage will win the space.
  - andymoe 11y ago
    Maybe it will be difficult to fully certify in the way you are thinking but the definition of "Commercial" is certainly going to be changing as rules for drones under 55lbs are finalized in the US as they already have been elsewhere. I think the space Airware will be able to compete successfully in with their current strategy will shrink as those rules get formalized and "hobbyist" platforms are used to do more and more real work. (Not that I don't wish them the best of luck - there are great folks over at Airware.)
  - apapli 11y ago
    it's just a matter of time before the open source platforms add in the features required for FAA etc approvals. All the other components (aside from flight controllers and chassis) are already OEM hobby items anyway, just photographed and put into a flashy brochure. It's not a big step to FAA approvals, despite the fanfare this seems to attract.
  - yo-code-sucks 11y ago
    Is that so? https://www.dronecode.org/ https://www.dronecode.org/
  - monomyth 11y ago
    Are you implying that 3DR drones use a different software?
  - kenrikm 11y ago
    There is no reason why it would be any more difficult to certify then Airware. The 3DR stuff is already more then capable todo the task and 3DR sells in the commercial space. Source: I've personally done multiple long range (many miles) autonomous missions and landed within 3ft of the target landing zone, I have hundreds of hours of flight time with many different platforms (Naze32, APM, Pixhawk, DJI).
- apapli 11y ago
  Agree. And a standard is already emerging for quadcopters with OpenPilot's CC3D platform. That's already pretty solid and has a very healthy release rate / dev community around it.
- foobarqux 11y ago
  I think Airware is going to pull a Meeraki and transition to focusing on the management portion of a solution.
slavik81 11y ago
Good luck. There's so much diversity in the equipment, payloads and use cases for these systems that it's incredibly difficult to standardize. Cutting integration costs has been a dream for many years, but previous standards introduced more complexity than they eliminated.
- spc476 11y ago
  Go back to the late 70s/early 80s and marvel at the incredible diversity in computers, most with unique operating systems. Standardiztion will happen eventually.
  - slavik81 11y ago
    It's a good point that we've seen this before. I'm under the impression that many early computer standards were essentially de facto standards caused by everyone trying to be compatible with a market leader. e.g. "Unix-like" or "IBM PC compatible". That's probably not the only way, though. I guess the real question is "how are successful standards created?"
dharma1 11y ago
nah. It will be ardupilot (or another open source flight controller) on linux, and because it's on linux, you have access to all the other open source goodies you need for processing, analytics, connectivity etc. It's already happening - check out http://emlid.com http://emlid.com or http://erlerobotics.com http://erlerobotics.com for boards you can use with Rpi2 or Beaglebone Black
- ericd 11y ago
  These look nice for consumer applications, but Big Enterprise Companies typically aren't looking for something thrown together using a RPi, and some of them are going to use a LOT of drones. The hobbyist market doesn't really seem significant compared to the number that will be used by utility, mining, telecom, and other companies with sprawling physical networks of assets.
  - dharma1 11y ago
    The point is not about Rpi, it was just an example. The powerful thing is that it's open source, very active project (check out https://github.com/diydrones/ardupilot https://github.com/diydrones/ardupilot), and runs (among other things) on standard Linux distributions. 3DR has received over $40m of funding so far, it is not in any way just a hobbyist platform and is aggressively going after industry verticals like farming. They manufacture and sell their own drones but there is also a growing number of other manufacturers building drones using ardupilot. That's not to say that Airware won't have customers, since drones in general are a fast growing sector and there's plenty of room for different businesses. But I doubt it will be the de facto operating system for drones
  - ObviousScience 11y ago
    I'd still expect that we'll see something like a real time OS microkernel written in hardened C or Ada/Spark or kin with a standard suite of drivers to control sensors, motors, etc. Most of those kind of things get written as open source projects, and then compiled in to specific bundles by middle men who get certified by insurance companies to reliably choose and compile such bundles. Big companies (eg, Boeing) will naturally still just use the open source microkernel tied together with a few in house drivers, since they have the capacity to deal with software certification themselves. Really, I wish hobbyist drones would start having more functional co-processors, because I want one where the flight computer and the user land computer... aren't the same core.
Htsthbjig 11y ago
I have more than 20 drones of different sizes, some of them are big. I am never going to buy a drone that is not mine(closed source) when I could buy machines that are totally under my control. Being closed source means the NSA could easily snoop what I do at every moment, in Europe, send the GPS coordinates, the telemetry, pictures and videos all the time without my consent. US companies are becoming increasingly spies of the US gobertment all around the world, specially people that worked closely with Uncle Sam, like those guys from Airware. I don't see Chinese or European companies buying into this, when they can tailor open source into their specific needs. We already do that, so it is not that complex thing to do. I see this article as PR.
stefanix 11y ago
"That kind of stovepipe way of doing things isn't going to scale," It's amazing how many times this argument is made by proprietary systems people. In their world the web should have never happened. The only chance they have is to use lots of VC money on certification bs that tries to lock out open source projects. I smell PR as well.

39 points by katm 11y ago

amelius 11y ago

It sounds like vaporware :)

sjtrny 11y ago
It is just a pr piece.
- evilolive 11y ago
  It is definitely PR. Airware has invested a lot more into doing its own hardware, its own possibly certifiable low level code than doing exciting things (sense & avoid etc). Heck, no amount of money will let you write a single line of code on their OS for drones currently. But is the PR going to work ? It seems easy enough to deter the "serious guys" doing some "serious commercial inspection" from using 3dr. The same way people use cisco equipment on their network because they're told that's what you do if you want peace of mind.

andymoe 11y ago

I don't think the operating system to connect the worlds drones is going to be closed source. 3DR and their open source base is going to give Airware a run for their money as the de facto standard "operating system for drones" in the consumer and more and more in the commercial space. In the meantime people like Drone Deploy and Skycatch will fill the gap when it comes to data collection and processing which is where the real business is long term. Drones themselves will continue to become a commodity. Also, it's worth remembering that a Chinese company, DJI, has about 80Pct market share in the US for consumer and prosumer drones. Lots of competition for everyone.

freeqaz 11y ago
3DR's software doesn't compete in the commercial space. It's a hobbyist platform and it will be much harder to certify by the FAA or other agencies. That is a huge portion of being successful in the space. Especially when you're doing things that have more risk, like photographing electrical infrastructure in a densely populated region. The FAA is going to be the agency that determines if you can do that, and the companies that has that advantage will win the space.
- andymoe 11y ago
  Maybe it will be difficult to fully certify in the way you are thinking but the definition of "Commercial" is certainly going to be changing as rules for drones under 55lbs are finalized in the US as they already have been elsewhere. I think the space Airware will be able to compete successfully in with their current strategy will shrink as those rules get formalized and "hobbyist" platforms are used to do more and more real work. (Not that I don't wish them the best of luck - there are great folks over at Airware.)
- apapli 11y ago
  it's just a matter of time before the open source platforms add in the features required for FAA etc approvals. All the other components (aside from flight controllers and chassis) are already OEM hobby items anyway, just photographed and put into a flashy brochure. It's not a big step to FAA approvals, despite the fanfare this seems to attract.
- yo-code-sucks 11y ago
  Is that so? https://www.dronecode.org/ https://www.dronecode.org/
- monomyth 11y ago
  Are you implying that 3DR drones use a different software?
- kenrikm 11y ago
  There is no reason why it would be any more difficult to certify then Airware. The 3DR stuff is already more then capable todo the task and 3DR sells in the commercial space. Source: I've personally done multiple long range (many miles) autonomous missions and landed within 3ft of the target landing zone, I have hundreds of hours of flight time with many different platforms (Naze32, APM, Pixhawk, DJI).
apapli 11y ago
Agree. And a standard is already emerging for quadcopters with OpenPilot's CC3D platform. That's already pretty solid and has a very healthy release rate / dev community around it.
foobarqux 11y ago
I think Airware is going to pull a Meeraki and transition to focusing on the management portion of a solution.

slavik81 11y ago

Good luck. There's so much diversity in the equipment, payloads and use cases for these systems that it's incredibly difficult to standardize. Cutting integration costs has been a dream for many years, but previous standards introduced more complexity than they eliminated.

spc476 11y ago
Go back to the late 70s/early 80s and marvel at the incredible diversity in computers, most with unique operating systems. Standardiztion will happen eventually.
- slavik81 11y ago
  It's a good point that we've seen this before. I'm under the impression that many early computer standards were essentially de facto standards caused by everyone trying to be compatible with a market leader. e.g. "Unix-like" or "IBM PC compatible". That's probably not the only way, though. I guess the real question is "how are successful standards created?"

dharma1 11y ago

nah. It will be ardupilot (or another open source flight controller) on linux, and because it's on linux, you have access to all the other open source goodies you need for processing, analytics, connectivity etc. It's already happening - check out http://emlid.com http://emlid.com or http://erlerobotics.com http://erlerobotics.com for boards you can use with Rpi2 or Beaglebone Black

ericd 11y ago
These look nice for consumer applications, but Big Enterprise Companies typically aren't looking for something thrown together using a RPi, and some of them are going to use a LOT of drones. The hobbyist market doesn't really seem significant compared to the number that will be used by utility, mining, telecom, and other companies with sprawling physical networks of assets.
- dharma1 11y ago
  The point is not about Rpi, it was just an example. The powerful thing is that it's open source, very active project (check out https://github.com/diydrones/ardupilot https://github.com/diydrones/ardupilot), and runs (among other things) on standard Linux distributions. 3DR has received over $40m of funding so far, it is not in any way just a hobbyist platform and is aggressively going after industry verticals like farming. They manufacture and sell their own drones but there is also a growing number of other manufacturers building drones using ardupilot. That's not to say that Airware won't have customers, since drones in general are a fast growing sector and there's plenty of room for different businesses. But I doubt it will be the de facto operating system for drones
- ObviousScience 11y ago
  I'd still expect that we'll see something like a real time OS microkernel written in hardened C or Ada/Spark or kin with a standard suite of drivers to control sensors, motors, etc. Most of those kind of things get written as open source projects, and then compiled in to specific bundles by middle men who get certified by insurance companies to reliably choose and compile such bundles. Big companies (eg, Boeing) will naturally still just use the open source microkernel tied together with a few in house drivers, since they have the capacity to deal with software certification themselves. Really, I wish hobbyist drones would start having more functional co-processors, because I want one where the flight computer and the user land computer... aren't the same core.

Htsthbjig 11y ago

I have more than 20 drones of different sizes, some of them are big. I am never going to buy a drone that is not mine(closed source) when I could buy machines that are totally under my control. Being closed source means the NSA could easily snoop what I do at every moment, in Europe, send the GPS coordinates, the telemetry, pictures and videos all the time without my consent. US companies are becoming increasingly spies of the US gobertment all around the world, specially people that worked closely with Uncle Sam, like those guys from Airware. I don't see Chinese or European companies buying into this, when they can tailor open source into their specific needs. We already do that, so it is not that complex thing to do. I see this article as PR.

stefanix 11y ago

"That kind of stovepipe way of doing things isn't going to scale," It's amazing how many times this argument is made by proprietary systems people. In their world the web should have never happened. The only chance they have is to use lots of VC money on certification bs that tries to lock out open source projects. I smell PR as well.