Its interesting that they are doing cool things but not quite breaking out commercially
But the other side of the coin is lots of money is being thrown at UAS in Europe and America really is beginning to fall behind.
Maybe ETH will put out their own or an enhanced opensource AP one day
Comments
@Dara I do admit to open nature of innovation. Have been around in open source pre kernel 1.0 release and having seen what has happened, have no doubts about success of open innovation. Unfortunately there are counter examples to it as well, I tend qoute "source fire" as I watched it very closely - a case of open source going closed. Then there are innumerable cases of open source projects dying out. The key ( heavy disclaimer :- afaik - this is my observation )for successful open source projects is a project leader who is respected in the developer community for technical expertise, with vision and good political skills ;).
I am sure there are many projects where a core of people would act as the leader. Or else someone from the developer community would step up to the plate and assume leadership.
What I am sure of is Chris has been around and has even watching trends and knows more than most of us ( defenitely more than me :)) about all this and would have a vision for diydrones.
@chris do u want to share a bit of ur vision with all of us. Without revealing too much, so that the nice surprises ( e.g. Like APM 2) still remain suprises...
'For example a mapping project requires aerial data acquisition,
ground control surveys and data processing.
Of all of these steps the briefest is taking
the photos and with fairly low rates in
the developed world for hiring a full scale it is hard to justify a UAV.'
I am a developer of Pteryx system for mapping and I assure you taking the photos
the correct way is exactly you need very specific UAV. The process takes around 2h
per 9km2 and is the least predictable part of the job, requiring highest quality equipment.
Job processing can take days, but then you are working at clean office.The idea 'I just make a photo of that field' works for amateurs only above their own field.
'Ahh...... 10K for autonomous foam wing ? Seriously? Us UAV nuts can do so much better for that price tag!'
No you can't. The guy that developed autopilot full time for Fly'nSense during 3 years of his PhD did notably better product than diydrones community during 3 years.
Just watch closely how integrated and ease to use it is, how many research application they implemented.
Ardupilot is nowhere near that quality and when it will be, F&S will be somewhere else. They are just better focused and regularly funded, swarm robotics is a way of hidden sponsoring clever people in reasonnably managed countries.
I agree Marc, I don't think the person that knows the killer app has bumped into sUAS at the minute. But they will soon and we will all say oh of course!
There are some large scale agriculture UAS trials going on in the UK. The operator is CAA licenced and able to fly BLOS. So also is another chap doing coastal flights.
You can grab lots of data from LOS though. Turning the collect into something meaningful is where the work is.
@David, the problem is that we are still talking about toys... If the demand is that the systems fulfil aeronautic requirements, we are no longer talking about 10k for a system. Regulations are not based on the applications but on safety (ok, they should be :-), so I can see the argument for line of sight, and that is and will be the most limiting factor. And please tell me the commercialy succesful applications in Canada and Australia. In Europe what I see is still 90% science...
@Robo: Yes, and that is not one of the expensive ones. You just have to calculate for the cost of an engeneering team and the low number in sales. Of course "we" can also built autonomous planes, but if you want to start to sell AND live of that, do the calculation. The customer wants a working product, so you have to do lots of testing... also time consuming and expensive.
And Mike, you are totally right. I just looked around for microlight planes because they are not really more expensive then a "good" UAS and much more versatile to use... Developing world is a good point, but also not from a commercial point of view, since there is no money to earn.
My conclusion: We still need to find THE application that justifies intense use of UAS...
Ahh...... 10K for autonomous foam wing ? Seriously? Us UAV nuts can do so much better for that price tag!
afaik, ( read huge disclaimer :- all my understanding is theortical) from what i gather this should be achievable on APM once we progress a bit more. I guess challenge would be having enough of power on CPU to spare to do all the extra computation needed to do flock guidance. BTW does anyone know is this a sort of mesh network (robust) that is being used to communicate between these nodes ?
I like how they shake them to activate them... I'll be surprised if the Ardupilot doesn't have the multi-vehicle coordination abilities shown in that video in its standard code base this time next year.
I think we know the problems of using sUAS commercially...
Usability (regulations), reliability/robustness/ease of use, and finally, the price!! The cost of the end product (in this case 3D data) has to be 1.lower then the benefit and 2.cheaper then competitors (e.g. manned AC).
Combined this is true for a very limited number of civil applications right now. So most things we see is research, because there the benefits of using UAS are much bigger.
In fact we need new ideas for applications that can only be done with UAS and everybody wants it :-)
History repeats itself. America had little interest in turbine engines or rockets while Germany surged ahead.