Quadcopter with Kinect doing altitude hold and obstacle avoidance

More from the Berkeley team we wrote about earlier, now with better documentation. From I Heart Robotics: "Patrick Bouffard from the Hybrid Systems Lab at UC Berkeley has done a great job documenting the ROS 3D Contest entry shown below. So it should be easy to reproduce the results if you are lucky enough to have access to a Kinect and an AscTec Pelican quadrotor."


They also note that Ascending Technologies is offering a 15,000 Euro sponsorship for a team doing UAV research. Proposals due by February 13th.


And I can't help but notice that there's a bit of Berkeley robotics cluster forming. Along with all the UC Berkeley research, there's DIY Drones, Berkeley Bionics (exoskeleton makers), 510 Systems (who made Google's robotic streetview cars) and the work at theLawrence Berkeley National Lab.

Views: 5940

Comment by Roberto Navoni on January 26, 2011 at 2:35pm

We are ready to join the contest ... :) We have multipilot 2.0 for stabilization and atom board to put on our quad ROS and kinect ... I'm yet doing some test on kinect , ROS and URBI too ... Who whant help me ?!?

This is my next step after MP2.0 ... We use Ubuntu X86 OS and an ATOM board onboard.

Comment by Tom on January 27, 2011 at 1:24am

I've been wondering whether it would be possible to reproduce the same with ArduCopter + BeagleBoard + BeagleROS + Kinect with the overall cost down to about 1000Eur instead of 7000Eur as with AscTec Pelican + Atom board.

Any one out there has already had some success using BeagleBoard + Kinect?

Comment by Roberto Navoni on January 27, 2011 at 9:21am

I think that is better to use Atom instead of Omap because ROS for Omap is not updated as Atom  x86 revision of code .. It's wrong ?



Comment by Luke Barnett on January 27, 2011 at 10:17pm

I'd agree with you Roberto. I've used an IGEPv2 (beagleboard clone) board before and it is painful having to recompile everything specifically for it. I eventually got an Atom COM which has definitely simplified development. With it I have implemented an interface to my Mikrokopter using the normally unused USB port, with the ATOM  running ROS on ubuntu. The development become easier because you can use a normal pc with ubuntu installed for the development and testing, then deploy the ROS packages onto the onboard ATOM and pretty much be guaranteed it will run.


The disadvantages are extra weight and cost, but you do get more processing power.


My project has stalled for now as I have realised how much of the vital code is missing from the Mikrokopter public repository so I am switching to the APM platform. Eventually I'd like to explore using the Kinect sensor, or preferably the Primesense Kit if I can get my hands on one.


I'm definitely looking forward to your Multipilot 2 board Roberto. 



Comment by Tom on January 28, 2011 at 1:39am

Don't get me wrong please, I wasn't trying to discourage people from taking part in the competition or joining Roberto. Just asking, if someone out there interfaced Kinect from BB-xM already. I'm also not questioning the possible ease of development that Atom is offering. But for many individuals the tenfold price difference means a lot when taking a possible crash under consideration.

@Luke: I think 1GHz processing power is enough already.

Comment by Luke Barnett on January 28, 2011 at 2:14am

Hi Tom,


Equally I didn't mean to discourage you, I just wanted to convey my experience with both. Also I think you'd be surprised at the cost of the COMs being sold now, you can get them for around 3 x the cost of a beagleboard so not as expensive as you might think, but still a quite a bit of money to put in the air.


I'd be interested to hear how you get on with your beagleboard, ROS and Kinect. 




Comment by Roberto Navoni on January 29, 2011 at 2:24pm

... I choose my x86 board to add to Ardupilot32 . Is developed in italy by Eurotech based on atom cpu. My idea is to develop a carrier board for AP32 that support connection with this atomboard

this is an a picture of that board.

Comment by Roberto Navoni on January 29, 2011 at 2:29pm

@Tom @Luke,

The price of Atom board is around 400 euro for a sample... instead of 140 .. so is not so different . But the software  is ready to use , so i think that the price that you save you need to spend in man hour to port the code on beagle board .. but the problem is not the first porting , but the continuous upgrade , patch ecc ecc .



Comment by Tom on January 30, 2011 at 3:33am

Ok, so this is something I didn't know. According to the pricing list I received from AscTec a couple of days ago Atom board costs 1200Eur, which is 1632$, which is more than 10 times more than the 149$ for BB-xM.

Is there a company which sales them any cheaper, or do I have to ask for sample at AscTec to buy Atom for the price you mentioned Roberto?


You need to be a member of DIY Drones to add comments!

Join DIY Drones


Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service