Automated Parts Counter Helps Build a Small Business

We love to see projects undertaken for the pure joy of building something new, but to be honest those builds are a dime a dozen around here. So when we see a great build that also aims to enhance productivity and push an entrepreneurial effort along, like this automated small parts counter, we sit up and take notice.

The necessity that birthed this invention is [Ryan Bates’] business of building DIY arcade game kits. The mini consoles seen in the video below are pretty slick, but kitting the nuts, bolts, spacers, and other bits together to ship out orders was an exercise in tedium. Sure, parts counting scales are a thing, but that’s hardly a walk-away solution. So with the help of some laser-cut gears and a couple of steppers, [Ryan] built a pretty capable little parts counter.

The interchangeable feed gears have holes sized to move specific parts up from a hopper to a chute. A photointerrupter counts the parts as they fall into plastic cups on an 8-position carousel, ready for bagging.Β [Ryan] also has a manual counter for wire crimp connectors that’s just begging to be automated, and we can see plenty of ways to leverage both solutions as he builds out his kitting system.

While we’veΒ seen more than a few candy sorting machines lately, it’s great to see someone building hardware to streamline the move from hobby to business like this. We’re looking forward to seeing where [Ryan] takes this from here.

24 thoughts on “Automated Parts Counter Helps Build a Small Business

    1. Ingenious. Great idea. Love it. How do you pick up screws and other stuff? How about a robot arm with machine vision? Now to look at your arcade machines. those look very cool. Brings back memories from playing too much PacMan while trying to mentally escape from too much engineering homework at UI Urbanana.

      1. The bottom of the bowl is slightly domed, and the whole bowl vibrates, it’s not actually rotating. The geometry of the ramp is very specific, and is usually custom designed for custom parts.

        1. Well you can have one with the ramp width large enough for the largest part, and just have different bolt on tippers, flippers and straighteners for different size and shape. Sometimes they’re not mechanical, but an air jet in a particular place.

          1. Not that easy. I used to work on these on a daily basis. Getting one to work right for a given part is something of an art. They are also pretty expensive. It takes a real touch to get the parts moving up the ramp correctly, and a small change in the part or the coating on the bowl can make or break the whole deal.
            We had one that the part was a rubber grommet that took a metal insert, and when the supervisor decided it would be better for ‘work flow’ to put the insert in before running them through the bowl, which fed a camera for part inspection, it very shorty stopped moving the parts up the ramp at all. Turned out the metal insert ruined the coating on the bowl and everything ground to a halt. They sent the $30k bowl out to be recoated, and production was down for almost two weeks.
            So, yeah, not that easy at all. Or cheap, for that matter.

      2. The motion of a vibrating feeder is a little more subtle than just a simple rotation/oscillation. A modified orbital sander or rock polisher really wouldn’t work properly as an actuator. Check out some images of how the flexural supports for vibrating bases for these bowls are designed and you’ll see that the bowls are designed to rotate as well as move slightly up/down vertically during the rotation movement. This vertical motion assists the slip-stick effect which is how the parts move. This same effect is used in linear feeding systems; proving that this effect isn’t actually centrifugal, but rather due to the inertia of the parts and the slight difference in static and kinetic surface friction.

  1. Clever solution. As author mentioned in his vid, my solution was counting scales. Had some tubes of ARM4s, 24b ADCs, and differential IAs left over from another project, so managers, as is generally the case, were too stupid to know the true material cost when the budget was presented to make 12 component counters. Full scale mass of each counting scale controlled by using discrete settings for bridges’ constant current source. The strain gage platform was from ebay/ali (it makes the purchasing dept crazy when we buy stuff via them, which is truly a joy), and the shaker pipes and gates done by the factory’s machine shop. Mostly used to count xfmr cores and bobbins, larger through-hole components, and fastener hardware. We make about 200 to 400 kits per hour, which is about right for our semi-custom production line.


  2. An all grown up version built a few years ago. The idea was spawned from me seeing a shell casing counter(?) that was posted on HaD 6 months or so previous to this machine getting built. I didn’t build it, I just suggested this to Chris when he mentioned needing to come up with a way to kit his projects.

  3. Industrial automation is an arena of engineering that is absolutely amazing to me. You REALLY have to think outside the freakin box to get some of that stuff working. Who are these people? Where do they get their ideas? I knows books exist with many ideas already in print, but it seems each time something new has to be pulled out of a hat that is verging on a magic trick.

  4. I would have loved having a similar machine on inventory days back when I worked in electronic manufacturing. Counting scales are great but you still need to count a base amount of your parts first, and the lighter the part the more you need of them for the final estimate to be semi-accurate.

    1. Everything you stated is correct, but these days a lot of counting scales have memories for (typically) 99 products, so you only have to do the reference sampling once for each item. Other enhancements available include barcode scanning (to recall an SKU), database integration for record-keeping purposes and/or when you have more items than the scale’s built-in memory can handle, container tares linked to SKUs, label and report printing etc.
      e.g. https://youtu.be/_s081QEHs_4

  5. The very last cup receives 9 nuts in the video instead of 8. I really like your build as well. Not sure the actually percent error in this. Could probably be solved with larger hole spacing / slower counting to make sure the sensor and arduino have time to sync up again.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s