Testing:

Final configuration and sensor testing before the presentation. Although we do not have full autonomy, we have sensor input from the light compass, ears, and whisker switches.

Final build with sensors:

Robot is mounted with all sensors (sound perception, proximity sensors, food and home whiskers)

Some pictures of klausbot near project finalisation.

Laser sensor

Quick note - the data sheet for the
ADNS 6030 laserstream sensor

which we are using to detect optical flow, and hence provide feedback on the crabs actual movement in the world...

R.I.P. KLAUSYellow

One of our gumstix boards has died. Yesterday at approx 5pm KLAUS yellow stopped responding, and now won't boot at all. Connecting via the serial board reveals that it never gets as far as the boot loader, and we subsequently (and painfully) noticed that an IC on the gumstix board was very hot. Searching the gumstix mailing list archives produced this posting that appears to describe the same fate that befell KLAUS yellow.

So, one left - KLAUS red is now in place and working well - lets hope he can carry the team to victory! (or at least on a random walk and then home...)

Yes...it's 4 AM!!!

Couldn't sleep. I put a bit of time into thinking about the mouse problem...

I found two related articles ( this one and this one ) which seem to be very relevant. I think fiddling with laser mice may be the wrong thing to do. There's an updated optical chip (ADNS 3060) which I think is what we need.

I had a look at the data sheet for this and took a peek at the forth code mentioned in those links. We may need to connect the chip to a PIC, then interface with the PIC via the serial ports on the robostix, or it may be possible to hook up the chips pretty much directly to the serial ports. We can look into this next week. I need to get down to B&Q to get one of those door peep hole things and we can try that parallel port thing that Kostas got working ages ago with the peep hole installed to see if it focuses nicely. Then we can connect the exact same circuit to the serial port on the robostix, port the visual basic code to C and we should be on the way.

Easier said than done of course....

Can you hear me?

Yesterday I started thinking about the ears for the threat localisation. The main idea is to amplify both ears. When the signal from one ear crosses a threshold value a latch is switched on and a pulse is generated whose width is determined by the time it takes for the sound to arrive to the 2nd ear. The pulse width is integrated and passed to the ADC. That should provide the time difference.