Controller Simulator

[Jay]

Does anyone know of an application that will let you send data from one computer to another simulating a rbot controller and a dashboard program?

[VanWEric]

For a while my team was hoping to emulate the parallax inc software on a palm pilot to code on it and then download it, but that never quite worked out, prolly due to the q3 mod we started on, but that also failed...
To make a long story medium length, Team250 formally offers to help. Email me @ vanweric@aol.com or e.vanwyk@team250.org. I am learning c++ now, and have learned a moderate deal of PBASIC, so I can be a good grunt level discover of many things.

[K-RAD]

How about writing a code interpretor for PBASIC? The only nontrivial part would be handling the input for the operator interface.

I believe I have some C++ code lying around for an interpretted language that could probably be easily modified for it.

Throw in some graphics and a flexible game engine and you'll have a full fledged simulation

[ChrisA]

i guess im going the cat get part way out of the bag here.

maybe not...

but i will tell you that there is a rumor that wildstang has created a couple of cool things this year that you may want to check out... especially if you want to create a controller simulator.

if ur wanting to find out more come check us out in the pits...
well be at the northwestern regional and also at nationals

[VanWEric]

Some one will have to remind me to go and check out all the other pits... Come find me at Team 250's pit thursday at the New Haven regional -- i should be the one commanding my lackeys*cough* minions to fix something. I may have something fun for everyone, or at least everyone can recite the entire Hackers movie backwards. Good luck to all.

[ChrisA]

im not so sure about the reciting of hackers backwards but we do have someone on our team that can recite it forwards...

(we found this out on our bus ride back from the great lakes regional)

[CJMJ]

I use a Basic Stamp (BS2) from Parallax to feed data to my Dashboard program. I have a simple PBasic application which continually streams 26 bytes to my computer's serial port. I usually just have each byte cycle from 0 to 254 (or 255 depending on the byte) in ascending or decending order so I can see the changes in the dashboard. I even use it to check out our robot's code before downloading it into the robot.

[ChrisA]

heres something to chew on before you come check out our custom circuit board (stang sense aka current sensing board)....

our drivers use a brand new color palm pilot (donated by palm) which we have hooked to the dashboard port on the oi (operator interface). we use this to display our data from stang sense and on another screen we have it printing out raw values for our joysticks and for our potentiometers which we use on our crab. while the raw data isnt too impressive (though its quite useful when we need to calibrate our pots) the color 'bar graphs' we use to display our stang sense data. as our drive motors draw more current the bars will grow larger and change from their 'happy' state of green to yellow then to orange and if we seem to be drawing way more current than we should be well see a nice red pigment and the bar will grow to full length (all the way across the screen). as of this moment we dont use any software as such to lower our current draw when it gets too high. if for some reason we start having a problem with accessive current draw we can potentially automatically lower the current with software so that we never stall (wouldnt that be cool). at this moment its up to our driver to glance at the palm every once in awhile to make sure hes not staying in the red for long periods of time.

well this is all great but you might be wondering why this is really useful to have a device showing us our current draw even if we cant really stop it unless our driver notices it and reacts to it. after matches we download the data to a laptop and watch it in close-to-realtime playback. having a video of the match playing next to the laptop we can watch for peaks and see what we are doing that is causing our motors to draw so much current. from that we can make changes in software as necessary so that current draw is minimal.

hopefully soon we will be using this data to theoretically stop the circuit breaker from ever tripping and also prevent any motors from stalling.

our team would be glad to use stang sense junior (our model version of the current sensing board) to look at current draw on other teams robots. in fact we got quite excited when we were allowed to use stang sense junior to current sense team 217 thunder chickens robot. if anyone is having problems with current draw they can feel free to talk to us.

i encourage anyone who is interested in learning more to come see us in the pits during motorola midwest regional or at nationals.

[kmcclary]

StangSense is definitely a VERY cool thing... <grin>

My suggestion to you would be to next figure out a way to add realtime timestamp data to BOTH your video and your datastream. (There are many ways to do that.)

Then, at a later time you could combine the two with a video editing station. First make a "realtime bargraph animation video" of the sense data (preserving the timestamp data). Next use the two timestamp data streams as an editing guide to combine the two videos onto a single videotape.

You'll now have a version of the robot video with animated data superimposed onto it you can copy and share with your members.

- Keith

[Greg McCoy]

Originally posted by ChrisA
our drivers use a brand new color palm pilot (donated by palm)

Do you think I can get Palm to donate me a new handheld?

[Dave Flowerday]

Keith, we're already one step ahead of you. As the data is recorded by the Palm it is timestamped. This is precisely the reason: we want to be able to correlate Dashboard data to a specific instant in a match. We also record the competition bit so we can tell the instant the match starts. This way it is extremely easy to watch the playback of a video of a match and begin StangSense data playback when the competition clock is started.

[kmcclary]

Originally posted by Dave Flowerday
Keith, we're already one step ahead of you. As the data is recorded by the Palm it is timestamped. [...and we ] also record the competition bit so we can tell the instant the match starts. This way it is extremely easy to watch the playback of a video of a match and begin StangSense data playback when the competition clock is started.

Great forward planning. (I assumed you were timestamping it, but had to ask... ) Recording the competition bit was a nice touch! (That's MUCH better than trying to sync the surge at power startup with the light on the robot coming on in the video! )

Have you considered ways of now superimposing the data as an NTSC graphic over the contest video? Lots of advantages: Viewable anywhere, no special equipment or software needed, and you can freeze or slow mo the video in either direction without losing data sync.

Here's an example of a superimposed visualization: Create small, animated "color pie meters" in each of the four corners of the video to represent the four drive wheel currents. Three (or four) laps in different colors for Full Scale. This would be easy to interpret out of the corner of your eye, even if each were tiny.

I've included a sample: Three laps (Grn/Yel/Red), FS=50 Amps. Indication = 43.75 Amps = 5/8 on the Red lap (overwriting Yellow) = 21/24th of FS = 87.5% FS.

- Keith

[VanWEric]

How does your "Stang sense" do its work? I may be a bit of a hardware newbie, but I don't quite get how you get the original data.

[Dave Flowerday]

Have you looked at my post in the Technical Discussion forum? Look for the discussion with the title "StangSense: Wildstang's Current Sensing System." There's a decent writeup there. If you've already read it, then if you could elaborate on your question a little bit more one of us will try to answer it as best we can.

[ChrisA]

haha ive managed to pull an electrical guy over to the software forum.... muahahahaha

1 down 4 more to go

i knew ud give in sooner or later dave