Question for Engineer or hobbyist

ok, going share situation in hopes out there can answer of questions.  i post on several boards.

here have:  two propellers 8 core cpus, 3 basicstampiis different speeds, 11 small robots, 1 arduino, 1 arduino mega, several single core computers , 1 dual core , 1 quad core. 1 laptop , 1 net-book. 1 robot head.  broadband internet wired , wireless, linksys wrt54gl router. have entry level experience on basic , c++.

i have mastered art of obstacle avoidance. but, when try more, gets slow becomes not practical.

i have 2 web-cams can recognize faces, tell are, track moving objects, voice recognized , can hear answer in human voice. can recognize colors , flesh tones.

there data base can interact , learn me speaking it.  i have not tried ocr yet, sure work.  i have not tried object recognition, sure work. have used chatterboxes , open-cv, , many more open source stuff.

now, problem is, of scattered out on many robots , computers. how 1 go tying of these items 1 machine?  i lost of how this.  where start?  i have deep desire experience ai in machine. not give orders or commands , communicate , have have it's own free agency.  to watch learn , grow , become more sum of it's parts.

oh, btw, wife says have no more money dump stupid project. if in shoes. start?

i hope there engineer out there or hobbyist can answer question me.

thanks!

moviemaker

yhmmc@yahoo.com

quote

oh, btw, wife says have no more money dump stupid project. if in shoes. start?

dtmfa (oops, sorry - channelling dan savage there moment)



seriously, arrange things robot build linked wirelessly backend controller; organise microcontrollers act semi-autonomous "agents" that, part, simple motor controllers , sensor readers, - if needed situation - can act autonomously main system if situation happening communications , such can't handle immediately. arrange @ least 1 of pc motherboards have used on-board robot well, perhaps local , smaller version of "learning database"; times when communications main controller , database either not possible or not needed.

build robot using whatever parts can scavenge; in fact, make challenge find parts need junk around house, neighbors junk, junk see on street, in garbage cans, wherever - don't pay (don't steal it, either!). amazed @ people throw away. more amazed @ people give asking (maybe know kids have outgrown powerwheels? awesome robot base, there!).

at rate, going need robot base bigger (likely) small robots; size of upended kitchen trash can (they make robot bodies, too!). go library (or purchase, if allowed wife), try find old 1980's books on robotics (if need list, pm me , send 1 archive once @ home) , such - many such old-school hobby robots built ton of junk , scrap.

mount robot head on interaction aspects, plus other stuff mentioned (controllers , motherboard).

on backend, arrange motherboards , machines cluster of sort; database oriented, since (or be) "brain" operations; far obstacle avoidance routines , such being slow, you'll have work more on them (but you'll have nice base work with). put web-cams in head or on bot somewhere.

i think if can start combine things, sell off can't use, put things in coherent whole, , if can't find right piece use or such, leave off until can; think - if can (draw sketches , such, test fit things together, etc) - , build "junk" have amassed, , working (even if not fully) - show making progress, may point wife, if sympathetic interests , hobbies loosen purse strings , let have money complete goals.

otherwise, need second (or primary) job, sell blood, donate sperm bank (not sure if pay it), open lemonade stand, etc - money need...

Arduino Forum > Forum 2005-2010 (read only) > Software > Development > Question for Engineer or hobbyist

arduino