I don't know if anyone posted about this before (I missed it if they did).It's on display at the CES in Las Vegas.It struck me, because when I first started playing, I wondered why there weren't any 802.11 telemetry/control models discussed on here. Now I find a quadricopter design, that looks incredibly refined.No price yet (Their website mentions a loaner and developer support for $1200 US)Via EngadgetParrot Website
One important issue I failed to mention, is when using 802.11b/g/n, there must be a access point or operate in ad-hoc mode... I'm wondering if it is a full access point or operate in ad-hoc mode...
I'm planning on doing a lot this myself, one of the BIGGEST issues I have is microchip muddled the gcc port so bad... so... I'm going to use a clean port and redo the API and use JTAG for debugging. So, an entire project all based on OpenSource using openOCD for the JTAG interface for debugging. I did a Port a long time ago for IDT ('334 MIPS32) so, it should be straight forward. http://www.schmalzhaus.com/UBW32/ there is a new one coming out at the end of Q1. It would be nice to use a ARM, but the only one I found was a TI part that has an external memory... anyway... see how this goes...
UBW32 (32 bit PIC32 based USB Bit Whacker) Project
Main website for UBW32 project - a PIC32 based UBW
deisgn
Tim, interesting idea. I couldn't find it on Sparkfun (the site is already crawling and its still a couple hours before Free Day!), could you link to it please? How does it work driver-wise, are there PIC drivers available or are you planning on hacking it yourself?
I'm going to try and use a PIC32 (UBW32 from spakrfun) Host USB to 802.11 dongle, this way can replace radio or use any usb dongle (that has open source support). That's the goal anyway...
Xander, it's a WBD-111 from Wiligear with an XR5 wireless module. By itself it weighs just under 80 g, around 100 g with the XR5 and by itself it draws around 1 W of power (something like 8 W on transmit). I have it running OpenWRT, a router linux distro.
The board itself lacks useful IO pins and ADCs, so I'm going to have it talk to an Arduino Mega over a serial. The Arduino will act as a low-level hardware liason, just forwarding sensor input to the ARM board and having the rest of its power for accurate PWM. The ARM will be doing all the heavy lifting (stabilization, Kalman, navigation, video transmission, ...).
I was originally going to fly just with the Arduino (it has enough firepower to handle it all hands-down), but I didn't wanna buy a radio set (I'm going for a more of a utilitarian, clumsy, safe vehicle than for an acrobatic flyer like an MK) and the idea of using an Xbee was soon abandoned (especially since here in EU we don't actually have any usable long-range bands). So I found this XR5 module and thought "hey, that would totally solve my problem". But it needed a miniPCI capable board, something an 8-bit arduino just can't be. And when I got my hands on a board that could, i thought that I just might use it for everything.
The board and the XR5 module will set you back about $100 each.
Yeah, I think I'd be using an external antenna...maybe even a separate router for better signal strength.
What routerboard are you using? I'd been happy with just a microcontroller on board up till now, but this has me rethinking that. If I can find a fairly cheap computer that can run wifi and talk directly to the sensors and escs it might be worth the extra weight and power drain.
This looks to be made of foam, appears to have brushed rather than brushless motors (and a gear for the prop drive) and just overall seems more like a "toy" like the walkera quadrocopters than the other commercial options out there like XUFO, MK, UAVP, etc.
Every time a commercial, polished multicopter shows up I'm both happy because it looks so cool and frightened, because all it takes is a lawtard from the company to write a bunch of patents and slam, here goes a hobby. Same thing that happened a couple years back with some old public domain image recognition algorithms, a company no one heard of before had them patented and then C&Ded all the projects using it.
Xander, the latency shouldn't be that bad, if it's possible to play Unreal on 802.11, then it'll be possible to fly a machine on it too :-) and with QoS limits on the video the latency should be even better. I'd be more concerned with the radiation pattern of most notebooks and the iPhone, even though they have a diversity system there's still a lot of dead zones.
I'm currently working on a multicopter myself, the only RF link is an 802.11a (5 GHz) module. And the flight computer is a routerboard running linux. So I hope that this Parrot deal relly works as advertised.
I'm thinking that the latency (response time) over wifi would kinda suck compared to traditional controllers. The protocol just has so much more overhead. Even compared to xbee. Would be great though to just have one digital connection for control, telemetry and video. I may have to give wifi a second look...
The mention a downward facing cam for stability. Which sounds like optical flow. I wonder if that could actually take over for accelerometers and/or gyros on board.
Comments
The board itself lacks useful IO pins and ADCs, so I'm going to have it talk to an Arduino Mega over a serial. The Arduino will act as a low-level hardware liason, just forwarding sensor input to the ARM board and having the rest of its power for accurate PWM. The ARM will be doing all the heavy lifting (stabilization, Kalman, navigation, video transmission, ...).
I was originally going to fly just with the Arduino (it has enough firepower to handle it all hands-down), but I didn't wanna buy a radio set (I'm going for a more of a utilitarian, clumsy, safe vehicle than for an acrobatic flyer like an MK) and the idea of using an Xbee was soon abandoned (especially since here in EU we don't actually have any usable long-range bands). So I found this XR5 module and thought "hey, that would totally solve my problem". But it needed a miniPCI capable board, something an 8-bit arduino just can't be. And when I got my hands on a board that could, i thought that I just might use it for everything.
The board and the XR5 module will set you back about $100 each.
Yeah, I think I'd be using an external antenna...maybe even a separate router for better signal strength.
What routerboard are you using? I'd been happy with just a microcontroller on board up till now, but this has me rethinking that. If I can find a fairly cheap computer that can run wifi and talk directly to the sensors and escs it might be worth the extra weight and power drain.
Just my $.02
-Jamie
Xander, the latency shouldn't be that bad, if it's possible to play Unreal on 802.11, then it'll be possible to fly a machine on it too :-) and with QoS limits on the video the latency should be even better. I'd be more concerned with the radiation pattern of most notebooks and the iPhone, even though they have a diversity system there's still a lot of dead zones.
I'm currently working on a multicopter myself, the only RF link is an 802.11a (5 GHz) module. And the flight computer is a routerboard running linux. So I hope that this Parrot deal relly works as advertised.
The mention a downward facing cam for stability. Which sounds like optical flow. I wonder if that could actually take over for accelerometers and/or gyros on board.
Hmm...lots to think about! =)
-
1
-
2
of 2 Next