Cartesian robot with sensing size and colour

Hello guys, i'm trying to make cartesian robot with sensing different size and colour like;

5x5 cm red box, 5x5 cm white box, 10x10 cm red box 10x10 cm white box

its like conveyor belt like i will try to build it on a specific rectangular base which has a gripper and grip boxes, sensing its colour and size and leave it to its specific area on base.

I made small project on PIC and microchips and basic language but its the first time i'm using Arduino so little help will be great! :)

I know that i will use;

  • Arduino UNO ???
  • step motors
  • motor drivers
  • mechanical frame and rectangular base
  • gripper? how should i choose it? vacuum or electronic?
  • which sensors for choosing?

So how can i do it? I mean how do i let it to choose boxes? which sensors do i use?

I really appreciate any help you can provide.

best regars

any help?

I don't understand the setup. Can you post a picture? This thing is lifting and shifting coloured boxes or it's putting things into coloured boxes?

It needs to recognise both colour and size, because some boxes are the same colour but different sizes? That is a difficult task. Most Arduinos can't do that. A Rasberry Pi will have enough visual processing to do that.

The choice of gripper depends primarily on what you are gripping and secondary is the power source. If you already have a source of vacuum then a vacuum gripper is pretty easy.

sorry for late response.

well our objective is to choose boxes which has different colour and size. Colours will be red and white or white and black. Sizes will be 5x5 cm and 10x10 cm.

The pictures are helpful but they raise a lot more questions:

How big is this? How far away does it need to 'see' boxes?

Are you moving the boxes or moving something into the boxes? You said "choose" but what does it mean to make a choice?

It would be more helpful if your answers were longer than the questions. We can't see what you're working on and we need more information to be able to provide helpful responses.

Its gonna be like 1 x 1 x 0.5 meters.

Firstly boxes are just waiting on the table at the spesific location. When we start the program, gripper will move the area where all boxes are waiting and grip box, determine what colour and what size it has and move it to another​ location which is also spesific location because all areas are written on the table and gripper just put box on the area. After gripper will move back to boxes areas and grip another one determine its colour and size and move it to their locations.

sorry for my bad english, i hope i made myself clear.

Do you know where the target areas are in advance or does the vision system have to read the table markings too?

This is starting to sound like a relatively advanced vision system. Beyond the capability of a plain Arduino. Maybe a Pixy camera can help but I suspect that you’ll be going for a high-end image-recognition library on a Rasberry.