GardenBot

One thing I noticed (and I am not too fond of) is the 'local circuit'. Having a 'modular' system that requires modification of the main control board (in my eyes) defeats the purpose of modules.

Perhaps it's an electronics restriction, but would it be possible to just create a complete module with ALL required circuitry and just run power and data lines back to the 'brain'? that way all that would be required would be software modifications.

I see pre-written code for various modules, and a table of what pins are what module, so in a main menu (lcd on the brain) would allow me to set which pins the new module is plugged into, and what type it is, which should allow very easy changes, and almost no new code would be required (except for new modules that don't have control code already pre-programmed). All configuration would be set via a text based menu (or web based, jumpers, etc). There's 1K of eeprom on-board so the 'map' of module->pins could be retained after a power loss.