Cool
Cool well whatever you think will work best 
Ok that makes sense
That makes sense. Ooh polymorph stuff?! Sounds pretty fun!
**Whats life if you cant have**<br><p>Whats life if you cant have a little laugh now and then
Ideally it could be adapted to any size trough just by adjusting the rail length. Alternatively the robot could be mobile and set free to search for plants that need tending. I think a university did that already with a bunch of roombas but I would like to give it a shot too.
Ive been having a read through the tutorial PDFs and just got through par, chan and select. Good lord why did it take me so long to discover parallel processing? Im sure there will be quite a learning curve but right now it looks like the holy grail for robotics. I can think of a hundred different concurrent threads I want to run!
yeah
just wanted to make sure i got a title, description, and such, to make it a proper entry =).
thanks
Yah, i normally work with the arduino and have had issues of wasting processing power on checking sensors and such taking away from what the main task is. The solution to this would be to use several low power processors (ugg no fun) or one big one (yay!). Here is a nice opportunity to do the second.
Welcome to the parallel XMOS world!
Welcome to the XMOS parallel world! A whole new set of possibilities are now open to you! I am glad it is all making sense to you now! Are you reading the XC tutorial PDF? That is indeed very nice, taking you through all the basics you need to adapt yourself into one mean parallel programmer
would the dev boards come
would the dev boards come with a programming cable, if not what kind is needed?
Lawn Care Bot
Project Title: Lawn Care Bot
Description: This robot will water your lawn, apply fertilizer, apply weed killer, apply grub preventer and apply lime. I plan on having 2 hoppers on the robot. 1 for water and the other for the lawn treatment. The robot would go to a central base where it would charge itself, fill the hoppers and be stored there until the next time water or treatment would be applied.
Kits: 1 kit
Diary: Photo journal and occasional videos of progress.
yeah
dont know ive never used it but i found it reasonably cheap so i bought it, it should turn up this week (hopefully)
wireless reminders
I find this pretty handy - I have a pretty fledgling system with a base station mounted on the door - it tells you if you’ve left windows open, and later I hope to get it to indicated if i’ve left lights on, air con on (or off). I’m using small wireless transmitter and receiver pairs. woo!
my idea
Seeing the power involved with this unit here is the idea that i have:
Title: Web enabled helper robot
description: A larger robot, in the 50 to 75 pound category, that has wireless video and audio, patched to internet via wireless router. The robot would be capable of voice recoginition, accepting commands verbally, and also totally web enabled. I can see the potential for handicapped individuals to answer doors, phones, and with the use of servos(large) and robotic arms, capable of doing smaller menial tasks.
The main thing is that it would need to be super user friendly, so that even inexperienced users could interact via an onboard keyboard, touchscreen, or any web enabled computer.
I envision the robot being able to read twitter feeds, as well as send them, check email and read the emails aloud.
I can see rfid enabled doorlocks that can only be opened by the proper pass card, or via the robotic interface.
When the user is away, the robot could be put into autonomous mode, and allowed to be a security guard. When an intruder is deteted, it could alert user via sms, twitter, or via post to a personal website. Possibly even via an outgoing phone call to 911.
Amount needed: I would think that possibly two or three would make this possible. Other parts needed would be various sensors, motors, lcd screen, and other various parts and pieces. A bridged internet connection can be done farily painlessly with the proper wireless router. Power could be taken from a gel cell system that would allow for automatic charging.
Log usage: well, i already keep a log of most of the projects that i do. Video is an issue right now, as a fall from the kitchen table ended my camcorders short life. Photos are not an issue, as I have multiple digital cams around.
My experience in programming is not that extensive, although I think i have a fairly good grasp of c, parallax’s basic(pbasic?), and also a lot of experience with Arduino’s ide. I have mainly speciallized in web site design and implementation, database design, and php, perl, cgi. I have limited experience with Ruby, flash, and vb.
I have no formal (School) training, as I started a family early, and have dedicated myself to raising kids, and paying bills. Most of the knowledge I have is totally self learned.
This is an ambitious project idea no doubt, but i figure if we are just spitballing ideas and possibly trying to build something new and different, why do it small. Make it big or don’t make it.
**oh **
oh can you ask your apps team ?
XMOS Sounds cool
These would make somethings really easy,
Optical flow navigation (memory management needs examination) but a couple of threads for nav and separate thread for biostatus, drive etc.
Subsumptive architecturres, easy one thread per node!
Sonar, DSP like functionality (Multiply Accumulate) means more information can be extracted from the return more than just time of flight, single thread per channel means things get interesting recovering 3d from the return signals.
Neural nets
I was going to look at the Propellers but XMOS sounds worth checking out, especially the XC compliler, C is my fav.
All the best
Cliff
Adding an ADC would be ok.
Adding an ADC would be ok. It is a very useful peripheral along with timers and communications methods. It looks like there are some good timing capabilites. Communcations would include SPI and I2C, those may have been already owrked out as well.
They should come with everything you need to get started
I have been discussing which kits would be best suited to all these projects etc with other XMOS people and if we decide on the one we are thinking of giving out, then yes, everything should be included to get started with programming the chip. You will need USB on your computer to do so.
Ooh another garden project
Cool looks like we have some green fingers in this forum!
very ambitious!
Well you certainly set the standards high there! As I have said elsewhere XMOS are providing some of our dev kits, though if you want to add other things like wireless/cameras/etc then this would have to be out of your own budget.
Glad to see a fellow web developer getting involved! Good stuff! I am also a web dev (as well as a zillion other things!)
I think my advice would be to tackle this project in very small steps and build it up block by block, or else it could get overwhelming pretty fast!
I am guessing this project might need something like 2 XC1’s and 1 XC2 (For ethernet connectivity - though you would have to write your own software that is capable of getting web pages / posting to them etc), I will have to see if this is do-able or not as that is quite a bit of kit!
XMOS Challenge
Hello XMOS,
Well I am a PIC and DSPIC user and a bit of an Arduino fan, but I have been looking at the Propeller as a solution to a few of the problems I am having.
1) Project Title: Telepresence Robot
2) Project Description
This project was inspired by the work the Leaf robot project; however I was unhappy with the platform as it is limited to flat areas. My robot is based on an articulated X platform (really H), normal operation is 4 wheels; however as each leg of the X is articulated and the backbone is also articulate the robot can stand on the rear wheels, and balance in this mode it operates as an inverted pendulum. Navigation of stairs and such obstacles is by alternatively standing on the front and rear wheels. The bulk of the mass will be on the back bone.<o:p></o:p>Processing is a mixture of Leaf code running on a Linux platform (micro ATX) booting from flash, I am thinking that the XMOS might form a good processor to do the sensor fusion of the low level sensors, Sonar etc, the other function will be motor control, steering and posture and balance, a little like the Amygdala/Hypothalamus/Thalamus system in the brain. I have been playing with subsumption on the Arduino but it not ideal.<o:p></o:p>Upper layers i.e. Navigation/decision/video being processed on the Linux board.<o:p></o:p>The head will house a Video camera and I suppose in a few years I will play with expression (I have a head from a Roboquad for this).<o:p></o:p>The role of the robot is to act as a Tele-presence free agent controlled from a 3G handset.<o:p></o:p>3) How many Dev Kits you think it may require (eg you may need 2 if you are demonstrating the ability of one robot using image recognition to track another for example)
1 Dev Kit, but I’d like a couple of loose processors if possible
4) Are you willing to keep a video/photo diary of your progress if we choose you? Please state which/both.
Yes, I will publish the CAD drawings and Source, this forum is about sharing the experence!
Thank your for reading this
Best regards
Cliff
Dev kit?
Am i correct the this is the dev kit you are talking about?
https://www.xmos.com/products/development-kits/xs1-g-development-kit
The list price for that kit is $999,- which is more than most robots on this site cost alltogether. I really like the specs of that device and I would definitely come up with some ideas for the XMOS if I had the time to actually implement them.
I don’t want to dis your product or anything and this promotional challenge is a great idea, but for $999 I think I would rather buy a netbook (or two) and use linux as a multithreaded environment with gigabytes of RAM and storage.
Yes we have modules that do that already I believe
Check out our community site for existing code related to SPI/I2C:
SPI:
http://www.xlinkers.org/node/47
I2C: