Back
[02:41:05] <rue_house> nope, rectangular
[04:28:27] <[VEGETA]> ?
[04:39:40] <[VEGETA]> <[VEGETA]> i have an idea for detecting moving obstacles... mere idea
[04:39:40] <[VEGETA]> <[VEGETA]> which is using laser scan data, by comparing previous scan with current one (frames)
[04:39:40] <[VEGETA]> <[VEGETA]> if there is a difference, then it is "dynamic" obstacle
[04:39:41] <[VEGETA]> <[VEGETA]> is it good enough?
[04:40:55] <rue_house> https://youtu.be/j4uz_yha9uI
[05:14:49] <Snert> rue_house: is the black/orange stripes on the base significant?
[05:17:02] <Snert> maybe that's how it detects which track the ball came down?
[05:28:23] <Jak_o_Shadows> I'd just have switches
[05:28:53] <Jak_o_Shadows> Whether those are pushbuttons or little light sensor things doesn't matter
[05:33:09] <Snert> I was thinking 5v on 1 side of the rail, ground in the other rail...let the ball make the contact and detect that.
[05:33:55] <Snert> or the ball completes a path to ground.
[05:34:05] <Snert> as opposed to messing with 5v.
[05:49:04] <Snert> maybe the express goal was to do it all visually.
[06:01:10] <Jak_o_Shadows> Hmm, that'd be a simple way of doing it.
[06:05:32] <theBear> https://www.youtube.com/watch?v=oHcH795M3qQ "standard robot protocol"
[08:02:48] <[VEGETA]> how to integrate a custom robot with ROS
[08:02:59] <[VEGETA]> mainly with navigation stack and similar packages
[08:03:15] <[VEGETA]> I don't think I can write a driver, that is why I am asking
[09:54:00] <[VEGETA]> back!
[09:55:06] <rue_house> Snert, its just decoration, the marbles are sensed by optoreflective sensors under the end of the track
[10:38:45] <[VEGETA]> looks like using a diff driver is easier:
http://wiki.ros.org/diff_drive_controller
[10:39:11] <[VEGETA]> you know a diff drive 4 wheel robot? cheap and can hold a kinect