Open source screw counting machine
github.com
github.com
I dump in a bunch of screws and bolts, and I know which are M4 x 3mm, which are 1/4" wood screws, etc.
If you have a bunch of small, used, jumbled-up fasteners, just chuck them in the bin. Small fasteners are cheap enough that it's not worth sorting them.
I have a storage bin rack on my garage wall, with one bin each for (new, used) * (nut, bolt) * (M4, M5, M6, M8, M10). (Actually, in practice, I don't have enough of anything except M6 to justify separating new and old, so only M6 is broken down that way).
But if you keep them sorted it saves a lot of time.
It depends if you count time in hours (you're mostly right) or days (you're wrong).
I have a bin of small screws and bolts which I've removed from devices. An hour spent searching for the right fit can save 1-3 days shipping time. If I'm in the middle of something, the advantage of being able to continue with whatever I'm repairing / building / doing ultimately saves hours of time in the end too. Breaking up a task each time I need a part loses focus and velocity.
What I'd really like is some machine vision system. If I remove fasteners from a laptop, I often don't know what they are. I'd like to be able to have the identified without having to look for service manuals. Ideally, it'd be attached to some way to sort too.
Totally agreed about how annoying it is to lose days to shipping time for trivial purchases though.
What does take up space is the part of your brain where you thought you had knowledge of every possible type of M3 screw/nut you could possibly need. After a few years of owning a "massive" collection of M3 screws/nuts that part of your brain will expand by an order of magnitude and suddenly you'll be looking to stock up on something like, "tapered torx head, alloy 18mm M3 screws" and maybe even nylon M3 screws/nuts.
Example: A few years ago I needed a square M3 nut to go into a 3D printed part. "No problem", I thought. I have plenty of square-shaped M3 nuts. Except the square nuts I had didn't fit in the part! "What the heck? Why is the slot for the square nut so thin‽" Turns out there's a standard for thin square nuts, DIN 562 where the thickness of the nut is 1.8mm instead of the usual 2.2mm and I didn't have any of those.
Now I have a little baggie of 50 M3 DIN 562 nuts (well, probably ~40 remaining) that is carefully marked separately from my other little baggie of regular M3 square nuts and I know better than to think, "I have all the types."
Hm, I didn't even know about the differences in square nuts, and I've been around the block a couple of times. Thank you for that new bit of knowledge. In return: setscrews... don't get me started on those. There are so many kinds, shapes and sizes that's a whole new category. And all manner of driver bits, I have so many of them I need to organize those too...
I bought screws from a company for several years when suddenly they changed the head by adding some knurls to M3 socket head screws. A different set of Truss head screws had a small countersink type feature added around the torx drive. And then the finishing process changed. All subtle changes but when one screw out of 20 is different it stands out. My solution was to track batches of cosmetic screws and compare them before transitioning to the next batch. Leftovers go in a spare parts bag.
While thinking about what to do I went down a rabbit hole learning about clocking screws. I decided that was far from necessary but if I ever make something really nice for myself in metal with the right design I might give it a try.
https://blog.lostartpress.com/2019/04/18/the-church-of-the-c...
Whoever worked on that church seems to have had the same idea :)
- Screw thread (M3 isn't the only type; and not all screws are ISO Metric)
- Screw length
- Intended material (wood, metal, plastic, matching washer, etc.)
- Head type (e.g. pan head, etc.)
- Driver type (e.g. Torx, Phillips, flat, hex, etc.)
- Screw material (e.g. brass, stainless, various coatings, etc.)
- Screw strength (e.g. structural screws which can handle shear, etc.)
- Special properties (e.g. self-tapping)
... and so on.
In most cases, there's a good-enough replacement in my big bin of oddball screws. Sometimes, I do need to redrill or retap the hole first.
Come to think of it, IF I had a quick way to identify them (e.g. a machine vision system), I could keep the organized by keeping DISsimilar screws together.
Sheldon Brown had a system where he kept 5mm/10mm/15mm/... in one color, 6mm/11mm/16mm/etc. in another. Color + very rough size was good enough to find the right tool: https://sheldonbrown.com/colorcode.html
Something similar might work for fasteners. Binning the would help a lot.
If I could identify them in the first place.
For which I need some kind of open-source machine vision system or something.
To answer your question: I have screws, bolts, and nuts left over from installing things, from fixing devices (e.g. laptops), etc. all sorts of place. This morning, I disassembled a clock which broke to see if I could fix it.
I tend to reuse parts when building things. The little plastic screws from the clock are perfectly usable. I can drill a hole in a piece of plastic, and screw them in. In this case, I don't need precision. For something like a laptop, I do. One of the bolts fell out of my laptop, and I'd love to be able to quickly find a replacement.
One thing that helps me when I'm disassembling something with an unfair number of fastener sizes is grabbing an egg carton, following a pattern I can replay, and using the carton to catch all the different sizes as they come out.
About 3 months ago I managed to pull that drawer out of the toolbox and spill everything over the garage floor. I wish I had a sorter for that day.
Have you done any redesigning to try and avoid a similar incident?
If you want to cover a wide variety of sizes including very similar sizes, then it could maybe become hard for a neural network to get the sizes right? So maybe just let the neural network classify the items and orientation, then measure dimensions in the image, maybe using stereo images. This will however now probably require manually labeling training images. Do you also want to read the grade and specification? Do you want to deal with coated, old, rusty or dirty items?
Interesting project, but that could take some time to get working reliable enough to be useful.
I can't stand the fact that screws and nails are measured by the pound (at least in the US). What a ridiculous unit of measure to use for a count of something. I have never, not once, thought to myself "I need 2 pounds of nails for this job" because only a lunatic would think of "how many" in terms of weight.
For very large numbers, it is always easier to count them by measuring either the total mass or the total volume of the items, instead of counting them, and then compute the numbers by dividing to the mass or volume of an individual item.
That is why in all chemical reactions the recipes are given by measuring the ingredients in grams or milliliters, even if what really matters are the numbers of molecules that must be combined. Because counting one by one the molecules is hopeless, the numbers of molecules are typically expressed as multiples of the Avogadro number, which is the number of molecules that would be contained in one gram of a substance whose molecular mass would be 1/12 of that of carbon 12.
So not only the nails, but also all chemical substances are usually counted by the gram, by the pound or by any other mass unit.
After all, it can be done with money…
If I need a handful of 5/16" screws, then odds are that I will also need nuts, washers and possibly lock washers. Instead of having to count out 10 each of all of that, I just grab a handful of all of them, toss them into a bag and the cashier weighs them and tells me the cost.
If I wanted a screw counting machine I would think about building this, but since I don't at the moment, I would like to know how it works.
If you are just curious about its operation, then I think you may be disappointed.
> Some screws
Ironic.