Hi everyone,

Thanks for your links and suggestions. I went to digikey and it looks like there are a few devices from Omnivision that should do what we want.

To answer your questions:

@Michael and Andy: Why 8051? because that's the processor that the students are familiar with already after taking the microprocessors course at my university and also because the robot platform that we will be using is based on a SiLab C8051F340 so that's what we have to play with. If it really turns out to be a no go getting the required information out of an image sensor on this hardware then we'll have to reconsider but that will cost money and we don't have a lot of that so I'd like to see what we can do with the hardware we have... We could build a daughter board for the robot platform that uses a different chip but I'd rather not if we don't have to.

@Andy, The flocking robots in Reading university's experiments (Kelly and Keating's paper) were based on 8MHz Z80 processors and used IR sensors to "see" each other and ultrasound to avoid obstacles. My problem with that experiment is that the robots had to communicate to tell each other where they were and what direction they were heading and to do this each robot was given its own identity (each transmitting IR on its own carrier frequency in a FDMA system) so the number of robots was limited to the number of channels in their multiplex (10 carrier frequencies). I am interested to see if we can make a system in which all the robots are identical and use no direct robot-robot communication so that the experiment is scalable and follows flocking behaviour in nature a little more closely.

Best wishes

Chris