This is an absolutely mind-bogglingly complex task that people are able to do easily. I don't expect to be able to build it anytime soon, but I'm very curious how it could even be done.
First, the task at a high level "I want a 10K resistor!" You might have it in bins, or binders, or baggies, or strewn across your desk, or, it doesn't matter. However it's organized, a few seconds later you have it. To return it, you have to pick it out and return it.
The doozy of a question is: "How can a system do this?"
This means 3 parts:
Catalogue. The system knows what it has and doesn't.
Contain. The system stores what it has and doesn't.
Retrive. When something it has is requested it can retrieve it.
Serve. The user can be served what the user has just asked for.
Accept. It can accept things from the user.
Identify. It can either identify what it has received from the user or must ask.
Catalogue. It must add it to its catalogue of things that it has
8 ) Store. It must store what it has just gotten.
Er, helluva a "3 parts".
Let's keep things simple and see if we could get this working for just resistors. How could anyone possibly do all of the above? What might such a system even look like?
One thing I could think of is something like this. Suppose it had thousands of plastic layers, like pages of a plastic binder. If it knew which layer something was in, perhaps the layers could be controlled in some automated way (P. 340), so that it opens to that page and the component there falls out.
What about the other parts? Any other way to do this?
NOTE: This is a totally out-there idea that I am not currently building. Maybe in 3 years. But I'd like to know the feasibility and how it could possibly be done.
EDIT: I don't mean to move the goalposts, but the reason I said "thousands of sheets of plastic" is so that there is not extra space around the components, and it can be basically small, tabletop size! I would like it to be a fast, small-sized retrieval system for a few hundred tiny parts. The inspiration was what I read about storing their components in 3-ring binders of baseball card holders (9 to a page) - surely the most clevwr solution would not be much larger than that. . But how to catalogue? RFID tags? Those don't work for little tiny smd's... Some kind of user input may be best, perhaps after weighing the component? But dwsoldered components wouldnt weigh the same after use... So many questions...
You are researching an inventory management and storage system. A database can be used to track recognized and available parts, nothing new so far. Physical handling of the parts can not be unified easily, each kind of package deserves a different machinery, and possibly storage compartments of different sizes. Recognition of labeled items or bags can be standardized, to some degree.
Such automated inventory management systems are in use since decades, so feasibility is out of question. A DIY approach is limited only by genie, time, space and money for the required machinery.
start with google.
do so some research
watch videos
teach yourself and educate yourself.
this forum is to help you get past problems you actually have, not hypotheticals and theoreticals.
post your code, what problem are you having this this ?
you are aware that in 3 years the whole arena of offerings will be different, right ?
to get you started google pick and place, start reading.
before too long, you will know more than anyone not in that industry.
but what you ask is silly simple.
you put a part number on a thing.
you type in the part number and the machine goes and gets it.
look at any lending library. card file shows location, books are bar coded.
look at any food store. parts are arranged in a specific order and location.
there are vending machines, you put money in, press a couple buttons to select location and you get a 3 Musketeers bar
since you control both, where they are located and part numbers, you can get started on this project today.
Three of my local Pharmacies have these machines already, every entertaining for kids and us older nerds.
Like a large cigarette machine, but has X-Y axis travelling collection bin with a lever on it.
The lever trips the appropriate inclined shute to dispense a box/bottle of your required medicine.
curious48:
...You might have it in bins, or binders, or baggies, or strewn across your desk, or, it doesn't matter...
The whole system is significant, but handling this part requires a sentient solution.
Item 1: Cataloguing and keeping it up to date is your first big project.
The pics of the pharmacy machine remind me of the old fashioned overhead cable systems that shops used to use to send cash from the counter to the cashier and return the change. Endless fun when I was younger. Only reason it is not fun now is because no shops have them.
this probably goes with his last thread about a place machine.
since the industry has solved every one of the problems. and all components come in rolls that fit into the machines the parts are labeled for machine orentation and such. the concept of this thread is without merit
just google pick and place.
to put order into a mash of parts strewn on a desk is silly when you can order parts ready for use in the machine.
dave-in-nj:
this probably goes with his last thread about a place machine.
since the industry has solved every one of the problems. and all components come in rolls that fit into the machines the parts are labeled for machine orentation and such. the concept of this thread is without merit
just google pick and place.
to put order into a mash of parts strewn on a desk is silly when you can order parts ready for use in the machine.
You are incorrect, this thread is about an intermediate solution between rows of shelves with bins, which I see in electronics youtube videos around people's workshops (or pick and place machines with reels of parts) , and a shoebox full of components that some hobbyists have, that they need to paw through to get to the component they're looking for.
Essentially I would like to automate this simple solution:
So you are quite wrong and this has nothing to do with magnification or helping hands. But it's not a bad guess
I can't see any correlation between Ring binders with zip-lock bags and "pick n place".
They are so incompatible mechanically.
IIRC a big problem with the helping-hands Thread was your lack of background knowledge about machines. While I and others here are happy to help you really need to do a huge amount of background reading before you even think about embarking on your projects.
Everytime I look at a complex machine, or something that looks complex to manufacture I remind myself that it represents the evolution of several generations of experimenting and learning. The ability to make complex machines does not materialize overnight. Was it Isaac Newton who said of himself that he "stood on the shoulders of giants"?
You can start as simple as with a catalog (database), where you enter a part number and it tells you e.g. "page 29, bag #5". Then you scroll your binder to page 29, and get the requested part from the 5th bag on that page.
Later on you can add a motor to scroll through the pages, and a hand that can grip a bag at the given position. And something that can handle multiple binders...
First, a bit of clarification. You can skip this bit:
So...just to clarify. Some people in this thread thought this has to do with a pick and place machine, perhaps because of the helping hands I discussed previously. That thread (which was actually about locking down a joint from moving) concerns a mechanical, hand-used machine - not an automated pick and place machine, a remote manipulator letting someone work with their hands, cheaply.
Although I mentioned the application, what I was looking for was very specific about locking a joint, and with help from this forum, I found it. Thanks for your help
So for an update on that, I have the mechanical solution I sought for my very specific issue, and I'll be ready to order parts and test them. Since most of the difficulty (as others here pointed out) is mechanical, I actually plan on writing up this whole process up as one of my threads on my board: http://curious.boards.net which is about all hobbyists interested in moving toward manufacturing, not just Arduino, but Raspberry Pi and other boards, and as appropriate mechanical engineering or anything else. So if you want to see the ongoing development of the mechanical remote manipulator based on a pantograph design, you should register on and check my board http://curious.boards.net regularly. Anyone who posts substantial, informative things (as most people here do, who have been amazingly helpful), and is nice and follows our policies can also discuss any on-topic subject there as well.
Back to this thread. You will notice that it's entirely unrelated, and is a Feasability question phrased "How would you"? I realize that complex inventory machines exist (and after all vending machines are an everyday example.) These are not the types of things that I have in mind.
I'd like to clarify that what I'm asking about has more to do with hobbyist organization of components that are being prototyped with. Something someone might use who has very little in inventory. Not a complicated solution.
Finally I want to point out that some of the references people have been giving are about systems that might cost $2000 to $10,000. I tried to find a list price, and looked through:
Well, let's see: "I amortized the cost of the robot over five years, including benefits, and determined the cost would be the same for buying the robot as for hiring a technician." So, a technician must go at the very, very lowest end for $20,000 fully loaded (this is a ridiculous underestimate), * 5 = $100,000. That's the kind of price that you can never find listed online.
So, at any rate, we are comparing with a $20,000+ system. Of course, it can count very small pills very accurately, and store a large number of them :). What it likely can't do is take them back from you.
This thread that we're talking about is very different, and is just at the idea stage. The reason I wrote that this is something I would do three years from now, is to emphasize to everyone that this is not something I was actually working on or doing. It's just something I'm looking at the feasability of, in line with this subforum
Specifically, the question is whether/how a simple binder full of baseball card holders (amazon link, no affiliation link) which is 100x for $15.60 and put into a $5.99 binder (again no affiliate link) with some masking tape to tape them shut and use as labels, plus of course 1000 antistatic pouches for $14.81 (no affiliate link) and maybe an antistatic wristband for good measure with all that plastic moving around for $0.96 (still no affiliation), could be automated to save the manual lookup time, while remaining in the realm of a hobbyist bin full of parts. Because the parts I just quoted are for 1,000 parts. What size machine will sort you into that many compartments? How much will it run you?
So as a very practical matter, there is some room here. This question concerns feasibility: how would you do it? Could you do it? Is there something closer to the binder full of 9-pocket plastic holders? Rather than toward the pharmacy machine that can be amortized over 5 years for as much as the fully loaded costs of employing a pharmacist for five years (including compensation, taxes, healthcare, and all employer contributions)? Something more in the $10-$90 of parts range? If so how would you do it?
Now. To get back to the what people have written so far:
DrDiettrich:
Physical handling of the parts can not be unified easily, each kind of package deserves a different machinery, and possibly storage compartments of different sizes.
I will disagree. I said let's start by limiting the components to resistors: I think a unified baseball-card-holder-in-binder solution can apply to all of the ones a hobbyist uses, from surface mount to normal resistors with prongs. So what if we just look at resistors?
A DIY approach is limited only by genie, time, space and money for the required machinery.
I would like it to not take much space, not cost much in components, and be able to be put together quickly.
TomGeorge:
Hi,
Three of my local Pharmacies have these machines already, every entertaining for kids and us older nerds.
Like a large cigarette machine, but has X-Y axis travelling collection bin with a lever on it.
The lever trips the appropriate inclined shute to dispense a box/bottle of your required medicine.
Tom...
Fantastic links and a viable automation solution if someone is willing to devote that much space. Likely also can be sold for several thousand. (Still worth it to EE's due to their time value.)
In my case I'm thinking of something much more lightweight: a binder full of (unpackaged) resistors in sleevies type thing.
lastchancename:
The whole system is significant, but handling this part requires a sentient solution.
Item 1: Cataloguing and keeping it up to date is your first big project.
RFID, manual entry, voice recognition (works fantastic today), scale for weighing, vision system (take pictures) perhaps combined with bar codes (stripes or 2D). Multimeter based sorting (just in time actual measurement, user attaches to leads yhen drops into storage slot). The only problem here is that I've just listed 7 mutually incompatible approaches... I want the best one...
dave-in-nj:
But order into a mash of parts strewn on a desk is silly when you can order parts ready for use in the machine.
There is no pick and place machine, this is for hobbyists with strewn desks. I am not sure where you get the idea of a machine from.
This thread is about an intermediate solution between rows of shelves with bins, which I see in electronics youtube videos around people's workshops, and a shoebox full of components that some hobbyists have, that they need to paw through to get to the component they're looking for.
Essentially I would like to automate this simple solution, or come up with an equally simple version.
Also, regarding what is cheap/expensive/available/unavailable, I thought I'd write the below as explicit clarification.
Just to put thie $100,000 into context: on my forum (curious.boards.net)and elsewhere I'm developing a program for broke-ass hobbyist niche manufacturers who would like to do a first run with an affordable combination Arduino/Microcontroller + ARM / Raspberry Pi-like board that costs them about $6-$7.50. I'd like to find about a thousand niche manufacturers who would contribute to a kickstarter campaign where they pledge about $5,000-$10,000 if in exchange they are able to receive 680-1694 units at around $5-$6-$7-$8 (as low as we can get it) so that they can put it in their enclosure or with any other small commodity parts they use, perhaps by hand, and bootstrap their first 680 orders. Broadcom will not even talk to someone trying to place a $5,000 order. So the ability to get 1,000 units is the difference between being able to make their first run of sales or not being able to make it. If the kickstarter (Indiegogo) is successful, the manufacturers (hobbyists) will have about a year to develop their products, while we work on turning the $5M-$8M into 1M units, to be given out to a thousand niche manufacturers. (Larger manufacturers will not be allowed.)
So while you might consider a $10,000 automated dispensing system to be "cheap" for an electrical engineer with 30 years of experience who is creating niche products by hand and charging $120/hour while doing so, in fact I compare it with the idea of saving up and investing $5K to get 680 boards that are certified and can be shipped to the customers in their own niches, which include requiring a full ARM computer and full arduino. By that standard, an automated shelving system is a horrendous waste of money for the hobbyist/new manufacturer.
They just need to go out and buy 500 baggies and a large rack for electronics. If you know of some niche where you can sell, but barely, then use your budget to place a very small minimum order of a thousand embeddable devices (or fewer), you might not be able to afford any of that on a "proper" dispensing machine. Yet time is money.
I want to help enable my fellow niche manufacturers. And actually, I hope to be one of them, not make money off of them - I have my own niche I'd like a thousand $5/$6 raspberry pi-like + arduino-like devices for that are tiny, supported, documented, legal to embed and sell, and ready to embed.
Embedded engineering designers have been in touch with me after my post on this subject. At least three of them, you can see one in the thread -reason78.
I've started preparing a LOT of kickstarter (indiegogo) materials, and am already confirmed on Indiegogo and can put campaigns up there. So, that is my longer-term plan if it materializes. In the meantime I want to establish my credibility by making extremely CHEAP versions of things that help out niche developers and people starting out. The first Indiegogo campaign I want to do is for the mechanical remote manipulator, i.e. a super-cheap help that people can use. You can see my threads at http://curious.boards.net about this (which will always be clearly marked), though I haven't talked about the remote manipulator yet, and I'll always be fully transparent about my motivations. I'm just curious and want to explore, and learn, and start shipping products.
So sorry about that rant, but it should place into context the reason I focused on a super-simple mechanical "toy" (as indeed pantographs today are "toys" - "Modern versions are sold as toys", according to Wikipedia at least), is that I wanted to lower my design risk. In my previous thread, I didn't ask for too much project feedback on the whole basic idea, but rather just on a specific mechanical fixture in it.
It's something simple that should work, and which I can use to start establishing my credibility filling and fulfilling. Hey, I'm new to all this and just learning Follow me at curious.boards.net where I'll be posting my progress, and of course you are welcome to the community (as long as you are nice, substantial, on-topic, and follow our guidelines). It's a bit wider than Arduino, and people here recommended I make my own forum for the niche that interests me.
Still, I also found arduino.cc to be a fantastic resource - which is why I'm asking here about how you guys think what I'm asking about could possbly be made.
Please note that this therad is not about something I'm actually working on at the moment.
I buy parts in larger quantities, so the parts in envelopes in a binder is too small.
I have 10-12 Really useful boxes, and put the envelopes of parts, tubes of parts, reels of parts, in those.
SMD resistors
thru hole resistors
same for caps
same for LEDs/diodes
thru hole IC sockets
power supply wallwarts
thru hole ICs not in long tubes
SMD ICs, transistors, regulators
other mechanical bits (switches, buttons, etc)
terminated wires
ribbon cable
I have divided drawers for header strips and female headers/crimp housings from Pololu
Bare PCBs
and some that are generally kits of boards we sell often
etc.
The boxes have a label on the end find one quickly, and they stack up nice on a 8 foot by 12 inch wide shelf above my workbench area.
There's sstill some pawing thru involved, but we're pretty good at find a part quick now that we're a little bit organized.
You can make a database on a PC that has details of all the parts you have and what "compartment" each of them is in. If you want, say, a 10k 0.25watt resistor you can look up the database and it will tell you which compartment it is in. That sort of database would be relatively easy to implement.
If you move beyond that to a machine that can locate the parts, pick one of them up and place it in your hand (or anywhere else) then you are talking about a pick 'n place machine.
In regular industrial electronics resistors are supplied on a tape in quantities of hundreds or thousands. They are never handled individually until the actual moment of placement in the circuit board. And i suspect the price of the machines would make the pharmacy machines look cheap.
My box of resistors has new ones with straight leads and new ones with twisted leads and previously used ones with shortened leads. When I try to find one it usually happens that 6 of them come out of the compartment at the same time because they are tangled together. No simple DIY machine will be able to pick resistors out of that mess.
If you seriously want to design a machine that can pick up resistors the very first thing you must define is how the resistors are to be stored so as to be accessible.
IMHO this is an interesting idea that would be exceptionally difficult (impossible?) for a DIY enthusiast with a workshop containing lathes and milling machines and who is very experienced in using them. The sort of person who would have no difficulty designing and making a small 8-day clock.
CrossRoads:
I buy parts in larger quantities, so the parts in envelopes in a binder is too small.
I have 10-12 Really useful boxes, and put the envelopes of parts, tubes of parts, reels of parts, in those.
SMD resistors
thru hole resistors
same for caps
same for LEDs/diodes
thru hole IC sockets
power supply wallwarts
thru hole ICs not in long tubes
SMD ICs, transistors, regulators
other mechanical bits (switches, buttons, etc)
terminated wires
ribbon cable
I have divided drawers for header strips and female headers/crimp housings from Pololu
Bare PCBs
and some that are generally kits of boards we sell often
etc.
The boxes have a label on the end find one quickly, and they stack up nice on a 8 foot by 12 inch wide shelf above my workbench area.
There's sstill some pawing thru involved, but we're pretty good at find a part quick now that we're a little bit organized.
I have four questions for you!
How many total (non-identical) components do you have?
What is median number of each (identical) component that you have? What is the most?
What is the size of the largest component you store in this way? What is a median (typical) size?
Suppose I asked for a particular component, let's say a 1K trimmer. How long in seconds would it take you to retrieve this?
Thanks for your answers to my questions. By the way, I think the setup you have described is extremely typical and many, many people seem to set up their work area in exactly this way (or an equivalent way) - as judged by youtube videos. Thanks for your answers and help!
Robin2:
My box of resistors has new ones with straight leads and new ones with twisted leads and previously used ones with shortened leads. When I try to find one it usually happens that 6 of them come out of the compartment at the same time because they are tangled together. No simple DIY machine will be able to pick resistors out of that mess.
Actually, it's easy. Make the same mess, but do it this way. Lay down a sheet of plastic, lay down one resistor. Lay down a sheet of plastic, lay down the next one on top or next to it. Repeat it with the 6. Now even though the 6 would normally be bunched together, this way if you lift the fourth sheet (vertically, holding all four corners), hold down the rest of the sheets and blow with a hairdrier, you would get the fourth resistor. The only difficulty in this case is knowing what was resting on top of the 4th sheet to begin with. There is no possible jumbling. Also this kind of goes without saying but it kind of helps is the plastic is antistatic
IMHO this is an interesting idea that would be exceptionally difficult (impossible?) for a DIY enthusiast with a workshop containing lathes and milling machines and who is very experienced in using them. The sort of person who would have no difficulty designing and making a small 8-day clock.
Thanks, I want to emphasize that I'm not building this. I'm just curious how anyone might possibly do so. For example, even if we assume there's a dirt-cheap, reliable, and easy way to lift page 346 vertically of 500 plastic sheets, so that a simple air puff can blow out whatever was laying on top of it, still, if the user gives the system a component something back, that they're now done with, how would the system know what component they just got? that component could have been one they "checked out" weeks ago, a simple resistor they had lying next to another one. How can the system know where to put it away?
I don't like this approach for two reasons: 1) it doesn't actually answer the question of how it works 2) even the parts that might work such as lifting four corners of a specified page while holding down the four corners of the next page - how would you even do that?
And the output selection is left completely unspecified. so, this is why it's a 2019 project
there is an old saying that crap in gets crap out.
if you were to spend a couple weeks cataloguing all my parts and listing them, then sorting them, then making them accessible for a machine,
that would be wonderful.
the amount of intelegence you would need to even begin such a project would require the combined data bases of every IC manufactuer, with archives of chips that are out of production, and matching photos so your vision system would work.
put it anotyher way, if you were to make a pick and place machine, a simple robot arm could access almost everything you have. a Swiss army knife of parts, tweezers, vacuum, grippers, whatever to pick them, should not cost more than $200.
pick, drop on the table, look. pick, flip over, look agin. move, orent as needed, pick and place.
as we move to a more computerized society, we are forced to stop being careless and sloppy.
the premise that only sloppy people will be your target misses the point that slopply people will eitehr buy new to avoid looking, or do not mind the neat adventure it is to search through the myriads of bins and parts.
in order to make use a seek and sort robot, I would have to bring boxes in from the garage, the shed, the attic, the two upsairs closets, the stuff under the couch, on the other side of the basement, not to mention the boxes of stuff at my feet.
for what ? to find a 39 cent TIP-120 ?.
when you read the links for a home brew pick and place that costs less than $100. your $10,000 number is just not in the same zip code.
I just got up, looking on my desk, 3 shelves, 2 boxes and a bench, counted 14 stepper motors.
NEMA23- NEMA17, smaller ones, I know of have a tiny one from a CD drrive upstairs, and a stepper with a optical encoder on the back for postioning.
I could do some similar searcing for LEDs, resistors, or hard drives (5 btw) relays, from 5v to 230v, from 1/10 amp to 30 amps, mechanical or SSR.
to catalog that, well, life is too short. and the fun of looking is lost.
the flip side is the guy who has everying in bins and racks
you cannot offer organization to someone who is organized.
also, if you are planning to manufacture, you buy what you need, then catalog it, then is is neat and ready for production.
I know the idea sounds great, but when you develop a 'business plan' and do some market research, and define your target audience, the project might not be attractive to others as much as it is to you.
