I am looking to automate one of my tedious business processes.
Here is how the process works (changing details slightly for this example)
1.) We carry 2000 different business cards.
2.) We receive orders from our customers, and they choose up to 25 different business cards at a time (only one of each card and sometimes they choose duplicates).
3.) Currently, we manually pick each order from small trays that contain all of the business cards (they are sorted by number. I.e. 141 = business card 141)
example. Customer A chooses 1 of # 25 card, 3 of #47 card, 1 of #78 card etc.
As you can see this takes us long time. I have been mulling over a solution and was hoping I could get some feedback suugestions from the experts on the forum.
1.) This is what I was thinking. Use an ALD5 robot and remove the gripper, replace with a suction cup/vacuum with an electronic valve that would provide enough suction to hold card while the arm lifts and moves it.
2.) (this is the tricky part) since there are 2,000 cards, I need to figure out a way to bring the cards close enough to the arm for the pick an place action. This is my idea:
A.) Create a ferris wheel like movement with 12 x 12 (or larger) sqaures that would rotate on some sort of track. This way the cards would remain level during circular motion. The RIOS program would know that sqaure 1 contains cards 1-10, sqaure 2, 11-20 etc. When the correct square comes around in the circle, the mechanism would stop and the arm would pick and move the card.
What do think of this idea? Do you have a better idea? The easier and less complicated the better (keeping in mind it takes up a lot of space to hold 2000 cards)
I’d say use a web page with the samples on it and let the customer click on what they want. Otherwise put barcodes on the samples and then feed them into a mechanical sorter/reader.
Might help to think in 3 dimensions. You could easily have 6 layers if you allowed for 2" between layers.
edit: heh, nevermind… allowing 4" x 2" for a slot I went the other way and said 48" seems a reasonable diameter… 48" x pi / 4" = 37 per layer, so 2000 / 37 * 2" / 12" = 9 feet high. and none of this considers how many cards are in a stack… is a stack of cards 1 foot long? 2 feet?
I don’t think the idea of stationary cards and moving pick is as space conserving as a moving carosel and stationary pick point.
First off, thank you for taking the time to respond, it is really helpful!
The cards will only need to be stacked up about 1 inch high (maybe 50 cards per slot). Multiple layers and a stationary pick point is exactly what I was thinking.
The question is, how far will the ALD5 reach? I was under the impression it was only about 12 inches, which means the layers themselves can not be to big.
Can you explain how the carosel would work? this is where I am having problem. What is the easiest way to get the layer in picking distance from the robotic arm?
An array of 45 x 25 bins of cards could conceivably be built into a 15’ X 10’ area. Then a router style (X Y) access to each of the bins could be created. Or break it up into several 4’ X 8’ bin sets. This is probably more do-able.
Instead of a router head, you’d have a pick-up (suction) device.
I thought about the x y axis solution, but the size of the robot would be extremely big and expensive. I still think a stationary pick point and moving trays/bins would work the best.
In this video the arm has a suction cup and picks and places in the box, which is exactly what I am looking for. The problem is that I need to create a device that would alternate the table top. Imagine a series of flat table top looking squares that would each contain around 20 different cards. The trays would rotate based on which tray was being called by the application.
I.e. tray 1 contains cards 1-20, tray 2 contains cards 21-40 etc.
My question is, what is the best way to bring the trays to the picking point? I was thinking of having stacked trays that would rotate up and then to the back (see attached file) for diagram.
how about a card shuffler / money counter type device? then use a conveyor with sections in it for each order. As it passes under the right card, each order/ count of cards drops to the belt in the right slot?
Build a vertical elevator box with slots to hold your trays (100!). Elevate to proper tray, then kick it out onto a table. Pick your cards, push try back in. That’s how a component tester project I worked on did it for LSI parts. But the “box” was only about 8" high, and held probably 12-20 2" x 2" LSI chip carriers.
You’d need to make your elevator MUCH bigger! Do the math on the tray size, and the spacing between each tray.
I’m a little confused with your idea… Where does the robot come into play here? j/k
Just out of curiosity here, you’ve been trying to keep the cards horizontally this whole time, but if the cards are in a case of some sort, couldn’t you place them vertically? That would allow you to place many more styles in one area.
I simply question whether a robot is the best solution to this particular business problem. And should the wider process be optimised.
Resolution of the pick problem will create new problems in loading the trays.
Given the availability of fast printing technology, it appears simpler and lower cost to use a print on demand
solution. This reduces the task compexity into a simple software problem. It removes the inventory management
cost, and the cost of inventory for 2000 sets of cards. The loading is of a blank cards only.
Use all the spare time and money you will then have to build a Hexapod or a MechWarrior.
he may be feeding the raw card stock, with pre-printed patterns on them, to a printer where the person’s name and stuff gets added. we really don’t know what the whole story is yet of course but when we order bussiness cards it’s like 500 or 1000 at a time, not 2 or 3 at a time. so I am guessing this is samples of different pre-printed card stocks being delivered to people to help in the decision making process of what to order in bulk. it would also follow that if selling high quality printed cards is your bussines model then digital printed cards would be of a lower quality which probably isn’t what you want to try and sell people with.
j/k