Brainbot - Navigation Chassis

[JonHylands]

So, I'm starting a new project at work - going to be doing some outdoor navigation research and development with Brainbot. The tracked chassis is not really suitable for that sort of thing, so we decided to mount the CPU/sensors on something a little more suitable for outdoor driving.



I went out and bought a 1/5 scale RC rock crawler chassis. Just for scale, those tires are 8" in diameter, and the whole thing is 30" long. We chose this chassis because it would have no issues going long distances at speed, and it would be able to handle curbs and grass and such with no issues. Being a rock crawler, it also goes slow enough that the sensors can keep up...

Right now, it has the quad core mini-itx mounted, and is being driven around with my Logitech wireless joystick. The motor driver (a Sabertooth 2x25) is controlled from one of my Bioloid I/O boards, which is connected to the bus. A program (written in Squeak Smalltalk) takes the joystick input, and sends speed and steering commands over the bus to the I/O board and the steering servos (AX-12's).

Here is a short video of it under remote control:



Next week I'll get the camera (Prosilica gigabit ethernet) mounted on a pan & tilt, mount the laser scanner, and I'll start writing a simple RoboMagellan-style navigation system. I'm also going to be picking up some stronger springs for the shocks, since we're a little on the heavy side for this chassis.

- Jon

[Adam]

That's sweet. I'm working on a 1/10 scale crawler-bot, but I never have time for it.

[Psykoman]

Awesome project. Can't wait to see more videos of it.

[JonHylands]

Here's another video, rolling around outdoors in the grass/snow.



- Jon

[Suicidal.Banana]

Hey Jon,

Cool update (already replyd that in blog actually) have you seen this guys work?
http://forums.trossenrobotics.com/showthread.php?t=3510

Maybe its too 'basic' (considering the hardware you have is butt-kicking 'strong') but it could be a nice little head start? personally im 'jealous' & blown away by the simple but perfect (and im sure hard to realize) idea, and that he actually got it working very decently too

Then again i cant even begin to imagine what ways you've come up with to make it able to navigate around on its own, gps, rangefinders, compasses, heck maybe you could even make some GSM-antenna trigonometry, or a combi of either (or all) of the above even?

Just wondering & letting you know of that guys visual-gate-system, sorry if im saying stupid things

[JonHylands]

Yeah, I've seen his stuff - its interesting, but not really what I'm working on.

Some of the other guys at the Brain Engineering Lab are working on the vision system, doing object recognition on a cluster with some pretty heavy-duty algorithms.

I'm going to be working on that part that handles mapping and navigation, using a hierarchical topological base mapping system I've been thinking about for a few years. I'll be using the camera for some things, but it will mainly be for path following. I'll be posting a lot more details here in the months to come, because all the work I'm doing right now will be open sourced.

- Jon

[Suicidal.Banana]

nice to hear that, im looking forward to seeing more of it good luck with the mapping

[JonHylands]

Here's a new picture of it, with the laser scanner and camera mounted. The camera is on a pan and tilt made up from printed parts and a couple AX-12 servos.



I'm in the process of building a pair of quadrature encoders for the wheels as well. Here's what the encoder disc looks like:



It is 2" in diameter, and has 16 tiny (1/16" diameter) rare earth magnets embedded into it.

Here's what it looks like mounted, with the encoder PCB mount also installed:



That brown stuff oozing out of the holes is support material, and it will dissolve away once I put the part in the dissolve tank for a few hours. The PCB (which screws to the two support posts) will have a couple hall effect sensors on it, lined up to provide a quadrature encoder signal.

Here's what the signal looks like from a single hall effect sensor, held in place by hand while spinning the motor at full speed:



Its a very nice clean signal, so I'm pretty happy with that. I won't get much in the way of resolution with this encoder, but its only going to be used for doing gross-level outdoor odometry.

Right now I'm finishing up the design of the encoder PCB, another PCB I'll use for the bus interface and switching power to the camera and laser scanner, and I'm about to start on designing a new power board for this robot. The power board will be very similar to the one I'm about to start selling, except it will have much larger MOSFETs to handle more current.

- Jon

[JonHylands]

So, yesterday I wanted to test the GPS, and I got a quick lesson in why rear steering is dangerous on fast moving, high center-of-gravity vehicles.

I think I'll add a toggle on my joystick to turn it on/off while I'll testing the chassis like this...



Fortunately, as I say in the video information, there wasn't any serious damage - just a couple scratches and a bent diode heat-sink.

I'm in the process of designing and printing a new roll-cage (which I was planning on doing all along).

- Jon

[JonHylands]

And here's the new cage design, which I am printing today (and probably tomorrow also - there's a lot of plastic in it):



This cage will also double as a carry handle.

I've also got a little ring cage designed for the camera in front.

- Jon