After the OSADL Lugano talk & demo on the portable RTAPI / BeagleBone 3D 
printer, Alex and me have decided to do another stunt, this time we shoot for 
the CeBit 2014 fair and the associated Linux conference track: 
http://www.pro-linux.de/news/1/20441/cebit-2014-call-for-papers-gestartet.html .

my goals for this event are:

- we plan to reach a wider audience, and the german CeBit is a good venue for 
that, in particular towards the german industrial base
- we specifically plan to show, and have ready by the time, a second demo which 
is clearly 'non-CNC', intended to address the wider scope for process 
automation which I feel entirely feasible once the realtime APIs are a bit more 
generalized and usable from non-NML clients. This is also a welcome time goal 
to finish a piece of work to the extent it's 'reliably demoable'.
- 'jump the hoop' on a Android-based UI working over the NML successor stack, 
with the intent of paving the way for post-PC UIs in a wider form.

My current plan is to take a typical Arduino-based self-balancing robot kit, 
rip out the Arduino, replace the RT control by a BB HAL/RTAPI setup, and have 
the whole thing remote controlled by an Android-based UI (wifi). Ideally we 
could control the robot by using tablet gestures like tilting.

For the hardware, I'm still looking for a kit along these lines: 
http://www.kickstarter.com/projects/tkjelectronics/balanduino-balancing-robot-kit
 - happy to take suggestions on something cheap and available now; we would 
need a few of these (Alex, me, maybe Charles can be dragged into helping with 
the accelerometer/gyro interfaces which would likely be SPI). Any suggestions 
welcome. I am specifically looking for a cheap kit, the reason is: in this case 
several folks might chip in and work on the same hw base; this would not be the 
case if several people start with parts from the junkbox. Also, the 
accelerometer and gyro should have separate assemblies (not everything on a 
single PCB).

For the software robot side, I'm assuming for now we can replicate the control 
algorithm(s) as a HAL setup, including hw interfaces, control loops, Kalman 
filter etc. After a cursory look I think that is feasible even if unchartered 
territory as far as HAL comps go. It is unclear atm if motion as it stands will 
be part of the demo, or if we need to cook up something more specific or adapt 
the code.

For the software control-side, this would amount to a Qt-based application a 
bit along the lines of gladevcp, plus the zeroMQ and protobuf stacks which 
should be readily available for that platform. I am betting on Alex's UI skills 
here but looking at his http://qremote.org/ site and application I think this 
will work out.

--

I'm writing this to tap into the community wisdom and let you know how what 
follows does fit into the overall plan, maybe somebody has applicable 
experience with hardware and software of that kind, and recommendations.

I am aware this is a rather ambitious plan, but then I'll take grandiose 
failure over not trying any day. We'll keep a fallback position with the 3D 
printer 'only' in case this turns out infeasible or not feasible in time, but 
there's a good chance it will work out, it's still 3 1/2 months to go. The next 
milestone is the paper deadline Jan 6, 2014 so we better know by then whether 
it's Plan A or Plan B.

--

As for the status of the NML-replacement work, I am right now molding an 
initial implementation of the Status Tracking Protocol into a library API, 
since this will be used by several status producers and consumers. Based on 
that the next step is to replace the EMC_STATUS usage in task and the Python 
linuxcnc module by this STP library API to get this communication path 
NML-free. Provided the ship is still afloat, then the other parts like the Tcl 
NML API, and then onto the queued command interface.  

Due to changes in my private life I am not as free to work on the theme as I 
used to, so any help and critical feedback would be welcome - sometimes I do 
get the impression I am talking against a wall of solid disinterest, maybe some 
not-invented-here here with rather few exceptions (thanks btw!). Not that this 
deters me, it's just a bit irritating to see this complete disinterest in a 
viable longer term evolution path for the project. But then the impression 
might be wrong, or folks just might have radically different ideas about what 
'development' and 'maintenance' actually mean.


- Michael


------------------------------------------------------------------------------
Rapidly troubleshoot problems before they affect your business. Most IT 
organizations don't have a clear picture of how application performance 
affects their revenue. With AppDynamics, you get 100% visibility into your 
Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro!
http://pubads.g.doubleclick.net/gampad/clk?id=84349351&iu=/4140/ostg.clktrk
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to