Here is a design for a light seeking symet that I have been working on. I was going to build wilf's bi-symet, but I couldn't get past only having two eyes. It really breaks up the symmetry of its namesake.
This design uses 3 LDR's with a set of spst mercury switches per 'side' for 6 total. I almost named this circuit 'toxic'. The tilt switches are oriented horizontally so only one set is activated depending on which outriggers the symet is resting on. The switches swap polarity on the corresponding LDR's, thereby changing the direction of the motor, and at the same time activate an optional fled indicator. The 100k resistor leaving the photobridge and the connection between pins 2 and 3 on the BBPV V2 create a latch so that once the photobridge is sampled, any changes in light will not affect the direction of motor movement for the duration of the t3se pop. The simple motor delay lets the photobridge settle before sampling and engaging the motor. The 1M resistor is connected from the output of the photobridge to + for good housekeeping in case somehow none of the switches are making contact. The 100 ohm resistors from the mercury switches are there to avoid a short if somehow multiple sides are making contact. With proper tilt switch orientation, this can be avoided, but better safe than sorry. Orient the tilt switches so that the contacts are towards the center, avoiding any potential for centrifugal force that might act in some situations.
I'm sure there is a less toxic, way to accomplish this behavior, but the symmetry is important to me and 6 mercury switches fit in there nicely. I have bread boarded this barring the mercury tilt switch logic, and it seems to work well. When I get my switches in from China, I'll finish this bot up and post some pics.
Have a look and any thoughts are appreciated.
Happy new year!