Quick help on proof of concept ?

Hi, anyone willing to hold my hand and help me build a simple project ? Its not that im lazy to learn its just my whole life i have struggled keeping up with technologies and now i have so much on my plate it would take me so much time before i can be able to get to my end goal and i dont plan to use arduino that much lot.

The project goal:

i want quickly see if i can sort coffee beans or garlic bulb peeled vs unpeeled. So the way this would work is i would have a small conveyor belt which moves garlic and the color sensor (TCS3200) would look at the bulb which passes next to it if it sees its yellow it would assumin its peeled if it sees any color other than yellow then it has to inform the soloneid which which release compress air which will immediately hit the bulb and thus thrw it out of the conveyor belt so in essence all non yellow color objects will be ejected and leaving only yellow color pass through and be collected at the end.

what i have:

GY-31 TCS230 TCS3200 color recognition sensor module

Bsp Air Speed Regulating Pneumatic Fitting Valve Throttle Valve Fast Connection M5/ 1/8" 1/4" 3/8"

NEW DC 4.5V 6V 12V 24V 2/3 way Gas Air Electric Solenoid Valve Electronic valve FA0520E

25W L298N Dual Channel H Bridge DC Stepper Motor Driver L298N Drive Controller Board Module 5V 2A Driver Module for Arduino

MEGA2560 MEGA 2560 R3 (ATmega2560-16AU CH340G) AVR USB board (lan) For Arduino

Small DC motors which can be sourced from CD-ROM drives / old radios etc etc

Let me know if anyone can help by creating code and visual directions. You can also PM me for skype chat if need be

Also see attached files for illustration

Thank You

Is this a comertial project?

zwieblum:
Is this a comertial project?

Maybe / Maybe not depending on how u look at it…im here in Africa I saw someone peeling garlic manually for a very small scale business (hand to mouth) now because it cumbersome i tried to help by devising a way to mass peel it works like 40% efficiency because garlic here is tough and much smaller but it speed up the process. but then on the other hand u end up with mixture of peeled vs unpeeled garlic so this brings another challenge ! i want to help and eliminate with the little resources and technical know-how

u can see attachment and realise what one has to go through to now sort it out !! so i got myself the arduino and thought i could easily try to see if it can help them

The garlic problem doesn't look too hard, but obviously the first thing is to see whether the sensor can reliably tell the difference between peeled and not. There's a tutorial here that looks useful.

ok, dark skinned garlic. Here it's almost always white. question is, are the pieces seperated far enough to do the airblow-thing? Or wouldn't it be easier to u use a rotating disk+trapdoor?

wildbill:
The garlic problem doesn't look too hard, but obviously the first thing is to see whether the sensor can reliably tell the difference between peeled and not. There's a tutorial here that looks useful.

Yes ... thanks alot for the link. will be useful atleast may reduce my learning curve

zwieblum:
ok, dark skinned garlic. Here it's almost always white. question is, are the pieces seperated far enough to do the airblow-thing? Or wouldn't it be easier to u use a rotating disk+trapdoor?

at the moment i havent thought of separating them far enough because if i separate them far enough then the rate of sorting will be extremely slow because we are sorting individual cloves. one whole garlic say has an average of 14 cloves/bulbs in one kilo we can have over 50 cloves remember these garlic are quite tinny not like the one we have in european supermarkets. in one Kilo we could have 700 cloves or more in a kilo and if they need to peel 50 kilos thats 35000 cloves atleast to sort !

tell me more abt this trapdoor concept im not familiar but sounds interesting ?

trapdoor thing: just a seperator that allows the stream of cloves to be directed to the left or the right - or to drop it in a bucket. it's easy to build (just washer motor + 2 endswitches), but you need some place to seperate the cloves, so that only one is affected by the trap. basicly you have a "portal" with the color picker, some cm behind it you have the trap. to guarantee, that the cloves arrive one by one seperated by some cm you can use 2 belts (slow+fast) or a birdfeeder arrangement (rotating disk, that places one seed in X seconds) + belt.

i was thinking of some sort of narrowing gate to a certain average dimension which would allow an average of only clove to pass as the conveyor moves in a particular direction

you'll have to guarantee that it never block. and you'll have to have all in on line, or the sensor won't see them.

Shaker bin feeding a drop tube. Sensor checks clove as it passes through an opening in the tube. Downstream pass/fail diverter (flipper, air blast, etc.) sends clove to appropriate collection point. Maybe a stop mechanism to prevent two cloves in quick succession passing the sensor.