I'm part of a student robotics club. While my knowlege of programming is near-nonexistent (I'm the "IT guy" - I just make the crappy computers we were given actually function), I'd like to see if I can use some scrounged parts and assorted bits I've got sitting around into a robot.
In the competitions, we've been using high-end PICmicros for "brains". That said, these are PICs - a TI-83 is remarkably fast when compared to a TI81, but it's still just a calculator. We have the same problem - trying to interface a video sensor has been insanely difficult, and autonomous operation is limited by the fact that our onboard CPU is, very simply, incapable of anything interesting.
However, I have a nifty toy I grabbed a while back - a single-board computer with a standard IDE controller, video output, and serial outputs. It's currently equipped with 64mb of RAM and a K6-II 233, but I hope to upgrade it to a K6-II 400 or better, and at least 128mb of RAM. It's only a bit bigger than a Nano-ITX board, and runs off of a convenient 12v. It might even have a USB port - my leg's broken, so I can't get the confounded thing out and check.
My plan goes as follows:
1. Build a control board, with a PICmicro, two ginormous H-bridges (yay, overkill!), rotary encoders, and hobby-servo PWM outputs. Using a serial I/O, this PIC will take serial data from the SBC, and use it to control the two drive motors and assorted hobbby servos connected. In addition, it will send back rotary encoder data, and can regulate the speed of the wheels by monitoring the rotary encoder. If programmed to do so, this should allow the robot to putter about - but it won't have enough power to do anything interesting.
2. Build a sensor board. This is, very simply, a PIC that takes data from assorted sensors, and sends it back to the SBC through a second serial connection. I'd also like to be able to use it to take data from other sensors that use serial connections, and relay it to the SBC.
3. Mount the SBC. Add assorted computer accessories, possibly including a webcam, for assorted purposes. It's not like there's a shortage of them.
In the competitions, we've been using high-end PICmicros for "brains". That said, these are PICs - a TI-83 is remarkably fast when compared to a TI81, but it's still just a calculator. We have the same problem - trying to interface a video sensor has been insanely difficult, and autonomous operation is limited by the fact that our onboard CPU is, very simply, incapable of anything interesting.
However, I have a nifty toy I grabbed a while back - a single-board computer with a standard IDE controller, video output, and serial outputs. It's currently equipped with 64mb of RAM and a K6-II 233, but I hope to upgrade it to a K6-II 400 or better, and at least 128mb of RAM. It's only a bit bigger than a Nano-ITX board, and runs off of a convenient 12v. It might even have a USB port - my leg's broken, so I can't get the confounded thing out and check.
My plan goes as follows:
1. Build a control board, with a PICmicro, two ginormous H-bridges (yay, overkill!), rotary encoders, and hobby-servo PWM outputs. Using a serial I/O, this PIC will take serial data from the SBC, and use it to control the two drive motors and assorted hobbby servos connected. In addition, it will send back rotary encoder data, and can regulate the speed of the wheels by monitoring the rotary encoder. If programmed to do so, this should allow the robot to putter about - but it won't have enough power to do anything interesting.
2. Build a sensor board. This is, very simply, a PIC that takes data from assorted sensors, and sends it back to the SBC through a second serial connection. I'd also like to be able to use it to take data from other sensors that use serial connections, and relay it to the SBC.
3. Mount the SBC. Add assorted computer accessories, possibly including a webcam, for assorted purposes. It's not like there's a shortage of them.