PDA

View Full Version : [Just For Fun] AX-12 Smart Arm Web Page Control Demo



MikeG
10-27-2008, 06:42 AM
I put up a web page (10/27/2008) that allows you to control an AX-12 Smart Arm remotely.
http://www.crustcrawler.com/products/smartarm/remote/ccvideo/index.php

Hopefully some of you can give it a test drive.

Eric
10-27-2008, 06:50 AM
Cool idea! Sorry man, I gave it a shot, installed the ActiveX control and got a pop-up error, LinkSY~ Cannot connect error.
It also crashes the reset of my IE7 sessions. (Windows XP, IE7, latest patches, updates)

4mem8
10-27-2008, 12:02 PM
MikeG, cool works great for me,Time lag, but that is to be expected, Good work.

Zenta
10-27-2008, 02:04 PM
Ah! Cool, it worked fine from Norway too! LOL

-Zenta

Adrenalynn
10-27-2008, 03:10 PM
It actually works pretty well. Latency isn't all that bad, video streamed well from the Linksys. I look forward to seeing more granularity of control at some later time. Nice work!

[IE 6.0.2900. Comcast Business-Class 16M/3M, Folsom CA. Quadcore 6600, Win2k3 Enterprise, 16GB, nVidia 9600GTR/OC]

Matt
10-27-2008, 03:48 PM
Awesome, Just worked fine for me in Firefox 2.0

Maybe in time upgrade to add blocks to move around or something.

Eric
10-27-2008, 04:14 PM
Tried it again... and YUP! works! Very cool!

DresnerRobotics
10-27-2008, 07:05 PM
Tried it again... and YUP! works! Very cool!

Wait.

You're in IT and you use IE?

Shun the nonbeliever! Shun!

Adrenalynn
10-27-2008, 08:22 PM
By my weblogs this month (about 13,000 unique visitors) only 17.94% are anything other than IE...

80/20 rule applies.

MikeG
10-27-2008, 08:28 PM
@All: Thanks for giving the arm a try. The arm received over 400 commands today. I had to reset it twice once because the arm was too close to the wall and once due to a temperature alarm. I’m pleasantly surprised with the performance of the AX-12s. The servos handled themselves better than I expected. I need to implement a soft reset though.

@Adrenalynn: My original implementation allowed users to read/write to all AX-12 control registers but I decided to limit the movements until I had a chance to do some stress testing.

So no comments on the AJAX instrument panel – motor feedback?

IE caused all kinds of headaches when it came to AJAX and HTTP GET. IE7 has an aggressive URL caching scheme which caused my AJAX controls to appear dead while Firefox worked like a champ. I had to make each GET unique for IE to work properly. I'm sure there is a more elegant solution but hey it works.

Adrenalynn
10-27-2008, 09:47 PM
I can appreciate that. I uttered quite the expletive when I noticed I was about to smack your lamp the first keypress... ;)

4mem8
10-28-2008, 12:05 AM
Mike, It worked sweet in FireFox 3.0.3

Modelmaker
10-28-2008, 01:52 PM
Just tried it out myself. That is awesome!!

it works great in Minefield (new firefox alpha build)

darkback2
10-29-2008, 07:49 AM
Two things,

First of all it works just fine in Safari from a mac...

Second...would you post a tutorial on how to set this whole thing up? Seams like it could be useful to a lot of people...

DB

Eric
10-29-2008, 08:10 AM
Very nice...
"So no comments on the AJAX instrument panel Ė motor feedback?"
I found that if you are looking for things to test out, you may need to be exact. Ask, would you test grabbing a block and moving it over here... That kind of directive. Left to their own devices, most will do things generically. (I face that situation big time)
Also ask for feedback on specific items... "your impressions of the Interface? Would you have any recommendations to improve it?" That kind of thing.
And did you REALLY think you would get away with that comment Tyberius. :happy:
Thems fighting words...
Actually, I'm a web application developer/project manager by trade, and I hear all the time... IE is always the starting point. As much as IE has been evil.. Lynn is right... it's used all the time.
Join the MS collective...Resistance is futile... you SHALL be assimilated :mad:

darkback2
10-29-2008, 08:24 AM
Ok...sorry,

I feel like the controls are a bit clunky in that the arm moves so far for each mouse click. If I were you I would consider making two graphs for positioning the arm. Maybe have it so that people can make a series of movements on the graph and then click play and the arm would run through them.

OK...the sequencing part is asking a bit much, but in the least a finer grain per click would be good. Also try to separate the X and y motion of the arm. Moving the arm down should not also move it forward...

I actually spent a lot of time trying to get vivians arm to work like that with not much avail. I had to settle for two modes. In one mode the arm moved up and down on a horizontal plain, in the second mode it moved horizontally. Seamed to work out pretty well...

I really hope this is the advice your looking for...and also, I hope you post a tutorial.

DB

Alex
10-29-2008, 08:29 AM
Awesome work Mike! It works great in Google Chrome too:)


Actually, I'm a web application developer/project manager by trade, and I hear all the time... IE is always the starting point. As much as IE has been evil.. Lynn is right... it's used all the time.
Join the MS collective...Resistance is futile... you SHALL be assimilated


Bah! The only time I use IE is to check to see if the page I just finished writing looks okay in it. Oh yeah, and then of course when I find a site that doesn't work correctly in any other browser than IE;) I've been a Firefox user since the beginning, although the memory leak problem that they STILL have not fixed in 3.0 has pissed me off to no end. Recently, I switched to Chrome, which I might say, has been fantastic, aside from Flash not rendering correctly and occasionally crashing everything. Funny thing, Google makes a browser that can't run their YouTube videos correctly??

Seriously though, Firefox holds a huge chunk of the browser market (http://www.w3schools.com/browsers/browsers_stats.asp). According to Google Analytics (awesome tool for web developers I might add, and FREE), over the course of the past year, in the TRC forums, there were more Firefox visitors than IE visitors. But, in our main site, it's just the opposite; IE blew the pants off of Firefox.

DresnerRobotics
10-29-2008, 10:35 AM
^ Yeah but in the IT world, most machines on the network run IE for simplicity. Easier to keep everything M$ driven for patching purposes and a lesser amount of variables when things go wrong.

That said- those in IT usually rock Mozilla and only have IE on there for testing purposes.

SHUUUUUUN!

(BTW Eric: I'm an MCSE/MCA, and survived 4 years of Intel Admin-Chaos, so I know the borg side all too well)

Adrenalynn
10-29-2008, 10:53 AM
>> in the TRC forums, there were more Firefox visitors than IE visitors. But, in our main site, it's just the opposite; IE blew the pants off of Firefox.

Totally OT, and I pre-apologize, but being the MathChick, I had to point out that the first sentence is how you lie with numbers. But you get to live today because you included the second. ;)

Sampling a geek-forum isn't a neutral sampling. Neither is sampling a geek-website, but as you can see from the site, it's less geek than the forum.

Alex
10-29-2008, 11:36 AM
Totally OT, and I pre-apologize, but being the MathChick, I had to point out that the first sentence is how you lie with numbers. But you get to live today because you included the second.

huh? Maybe I'm misunderstanding what you're talking about, but I'm totally confused here Adrenalynn... Care to elaborate on what you mean by "how you lie with numbers"? I was merely pointing out facts about our site's statistics, both in the forums and the main site.

Adrenalynn
10-29-2008, 11:46 AM
It should have read "how one can lie with numbers".

If I only sample the log files from Mozilla.org, I'd probably see like 90% Firefox, 5% Safari, 2% Opera, and 3% IE.
If I only sample the log files for the internal MS site, I'd likely see 95% IE, <1% Safari, <1% Opera, and the remainder FF.
If I sample a geek.Forum with the majority being repeat traffic, Firefox will rank high. If I sample __insert generic website__, we'll see that Firefox is in the minority by a landslide amount.

The favorite trick of pollsters is to carefully select their demographic to get the answer they want, and then word the questions in a leading sort of way just to seal the deal.

Sorry, I thought the "But you get to live today because you included the second." statement would let you know I wasn't picking on you directly because you offered-up the second half of the equation.

4mem8
10-29-2008, 12:20 PM
Man, you have a way with words Adrenalynn. I usually have to read your posts twice to make sure I understood it, ha ha.

Alex
10-29-2008, 12:20 PM
It should have read "how one can lie with numbers".

Ah, gotcha, thanks for the clarification. Damn Internet and misinterpreting what people mean;)



If I only sample the log files from Mozilla.org, I'd probably see like 90% Firefox, 5% Safari, 2% Opera, and 3% IE.
If I only sample the log files for the internal MS site, I'd likely see 95% IE, <1% Safari, <1% Opera, and the remainder FF.
If I sample a geek.Forum with the majority being repeat traffic, Firefox will rank high. If I sample __insert generic website__, we'll see that Firefox is in the minority by a landslide amount.




Hence why my original post linked to Browser Stats from w3Schools:

http://www.w3schools.com/browsers/browsers_stats.asp

:p

agavejoe
10-29-2008, 10:52 PM
@darkback2: The "clunky" control is on purpose. I wanted make sure that the AX-12s would hold up and I wanted to see how the XML service handled the load. I instrumented the app so I've been making adjustments everyday.

Later I'll show how to control the arm anyway you like. Actually you have the ability right now I just didn't publish how.

darkback2
10-29-2008, 11:00 PM
Awesome. Please post directions on how to do this. It would be really cool to have a network of arms that people could log on and control from all over the world. I would put one in my classroom. The students would get a kick out of seeing the arm start moving...

It would be especially cool because then parents could log on and "reach out and touch" their children when they misbehave. I could just see a robot arm wielding a belt, as a speaker screamed "billy...what did I tell you about using your cellphone to text in class!"

DB

MikeG
10-30-2008, 07:50 AM
@darkback2: Like I mentioned, I’m simply testing the arm in a controlled environment. I finally feel comfortable leaving the arm unattended for several hours at a time. Before I’ll feel comfortable releasing full control, I’ll need to add a considerable amount of validation. When I say full control I really mean full control no UI except for the Instrument Panel and video. You can think of it as a distributed application. I’ll supply the interface you supply the logic.

darkback2
10-30-2008, 08:16 AM
@darkback2: Like I mentioned, Iím simply testing the arm in a controlled environment. I finally feel comfortable leaving the arm unattended for several hours at a time. Before Iíll feel comfortable releasing full control, Iíll need to add a considerable amount of validation. When I say full control I really mean full control no UI except for the Instrument Panel and video. You can think of it as a distributed application. Iíll supply the interface you supply the logic.

Sorry..."by how to do this" I mean how to set up an arm to be controlled over the internet. You've inspired me. I want to make one of my own.

DB

MikeG
10-30-2008, 10:10 PM
Overview:
http://www.crustcrawler.com/products/smartarm/remote/walkthrough.php

Adrenalynn
10-30-2008, 10:58 PM
Nice, Mike! It's a good start on any larger project like what DB is suggesting. Thanks for sharing (and continuing its development)!