Automated Pocky Dispenser

Sometimes, along comes a build project that is not so much a fail, as how not to do it. First off, some of us here had to look up what a Pocky is, never having heard, seen or tasted one – seriously. Once satisfied, we turned our attention to [Michael]’s Automated Pocky Dispenser. Took a while for us to figure out if it’s useful or not. But it’s a fun, quick project that [Michael] put together in around an hour using parts lying around in his office.

For those of you who’d like to know, a Pocky is a chocolate-coated biscuit stick, although you can also buy it in other flavors. You can grab one from a box, but maybe it tastes better when you dispense it by banging a big red button. [Michael] says he used  incredibly advanced construction techniques, but we leave it to our readers to decide on that. The key element of the build is the special “flexible coupling” that he built to transfer the rotation of the stepper motor to the dispensing mechanism. The rest of the build consists of an Arduino, stepper motor, driver, and giant red button. Special motor driving code ensures that the dispenser wiggles back and forth every time, preventing any stuck Pocky’s. And the Electronics are, well, hanging out for all to see. Happy with the success of his build, [Michael] is planning an upgraded version – to connect the Pocky Dispenser to the cloud for statistical gathering of office Pocky habits. He claims even Google does not have that data. To see the dispenser in action, check out the video below.

[vimeo 120897308]

36 thoughts on “Automated Pocky Dispenser

  1. Pockys look to be well suited to a launcher of some kind. Maybe a pocky crossbow or pneumatic system with a pan-tilt mount and camera. When I want one I would remotely control it and launch one at myself from across the office.

  2. I’ve been using thick walled PVC tubing intended for fuel line as flexible couplers with pretty good results. I get the 4mm inner diameter and warm it up before pushing it on the 5mm stepper shafts. Obviously used for light-duty stuff like this.

    For a project like this, I would have used a servo though.

  3. Mmmm, Pocky.

    Reminds me of all of the great junk food that I bought from a local ‘combini’ (convenience store) when I visited Tokyo.

    Wasabi-coated rice crackers, anyone?

  4. Arduino = overkill you could do this with a simple analog circuit and not even use a steeper. use a notch on the rod to detect when a rotation has occurred, use a regular motor and latching circuit.

    Press button, rotation begins a switch on the shaft closes, as it rotates, the product is dispensed and then the switch opens at the end of the rotation setting the system up for another pass. One push button, motor, roller switch and a battery = dispenser upgrade.

    They use the same dam thing on toothpick dispensers and you can save your self a lot of money in the build. No stepper, no motor controller, and no microcontroller & programmer (hidden cost).

    1. Why does it matter in the slightest whether this is the “overkill” way of doing this? Why is this such a constant problem on Hack-A-Day? No one cares.

      This guy built something. You’re sitting at home armchair-engineering it on the internet. If you want to complain about inefficiency, at least put your money where your mouth is and make one of these demonstrating your own better strategy.

      I’ll beat you for efficiency every time with my own design though: you put the box of candy on the cable, open the top, and people pull one out every time they want one. Total cost $0.00, total power consumption 0W, total code required 0 lines. Obviously this is what should REALLY be featured on this website!

      1. People care, it’s just that the ones who would post comments have been driven away.

        Seriously, for the popularity HAD has, it has *very* few comments, and the majority of those are not worth reading.

        The problem is that they (the site managers) don’t take criticism in a positive light, don’t act on suggestions for improvement, and berate the readers as a response to the problems – instead of fixing the system so that there wouldn’t be problems in the first place.

        Other sites have mechanisms that encourage insightful responses and push down trollish or boorish behaviour.

        Don’t bother responding to posters, don’t bother suggesting improvements, and don’t bother complaining.

        It won’t help.

        1. This is a response that plays directly into your assessment of the site managers, but I’m going to entertain myself for just a few minutes to parse and respond to this comment.

          > The problem is that they (the site managers) don’t take criticism in a positive light, don’t act on suggestions for improvement,

          I would like to prove you wrong! Surely there is a suggestion for improvement in this comment!

          > instead of fixing the system so that there wouldn’t be problems in the first place.

          Never mind, I stand corrected.

          And thus we come to the crux of your argument. You claim that I don’t respond to criticism, and don’t implement suggested improvements. You offer no criticism, and suggest no improvements. Seriously. When you sit down and actually read all the comments, you’ll find that most of them consist of, “Man hackaday has really gone downhill”, and “this was a terrible article.” Explain how I can improve on that.

          I mean that in the most non-condescending tone as possible. How, exactly, would you improve when your only advice to improve is, “you should stop sucking”?

          I never see a comment that says, “you guys should do more original content”, or, “focus more on domain x,” or, “there are too many projects involving x”. It’s always, “man, hackaday has really gone downhill.” Thank you for that bit of actionable feedback. If you’d like to expand on this, there’s a reply button to this comment

          > Seriously, for the popularity HAD has, it has *very* few comments, and the majority of those are not worth reading.
          > Other sites have mechanisms that encourage insightful responses and push down trollish or boorish behaviour.

          I would argue against the ‘very few comments’ assertion. Here’s EETime’s post on the nxp merger. Two comments. As of right now, Hackaday’s post on the same subject has six. That’s after half an hour. The number of comments on Make are usually in the single digits. Adafruit blog comments? Same thing. The problem is you’re comparing the number of Hackaday comments on any post versus the number of comments on Reddit, HN, or Slashdot. Those are different formats, and I reject the assertion that we get very few comments. We get a large amount of comments considering our format. But I digress…

          You’re referring to karma or upvotes, or some kind of user-based comment rating, right? That’s not going to happen. That’s the surest way to decrease the ratio of signal to noise and turn the comment section into a pile of shitty puns. I’ll go on record and say the comment system needs an edit button, and it should have greater depth. It does not need a way to upvote comments. You are, of course, free to argue that.

          I would like to note that you’re complaining about Hackaday in reply to someone that’s complaining about a comment, in reply to someone that’s complaining about an Arduino. You’re three or four levels away from being anywhere remotely on topic, and you want a comment system where people can hide this obviously off-topic comment? Think about what you wish for when you commit yourself to the will of the masses. At the very least, this system allows your complaints to be seen.

          1. I don’t think a upvote/downvote system is a good idea but a way of indicating someone is incorrect, off topic, troll, etc. has it’s merits.

            Threading is also somewhat of a mess (I have no clue if this will be threaded under you or show up somewhere else) Auto collapse would help. A way of seeing new posts would help (I often end up reading the same comments twice because I don’t know which ones I have read and which ones I have not – this is somewhat related to the threading not working correctly)

            My other issue is actually finding a old article can be tricky. You guys have categories, but they could be shown better. Perhaps a page with the list of categories, with one or two examples of the latest posts, from there more specific stuff. For instance you have digital audio hacks, but when I click that I get anything from synthizers to class D amps. That’s great but not too useful when i’m looking for one or two things.

            A better way would be:
            – Audio
            —Class D

            You could also organize some of the more reference stuff you guys have wrote like the 3d printer article. That should be a top link “Hackaday Tutorials” I really like most of the stuff you guys produce but you really don’t index it well. Same with your weekly articles like ‘the hacklet’ I can go but that really should be described under a more clear hierarchy.

            The site is also somewhat cluttered. The sidebar is far too long and all the useful information is at the bottom meaning I have to scroll down. It is basically unreadable on mobile (the right sidebar uses too much screen space, it should just be gone by default on mobile) Comments are completely impossible to read because they are all squished to the side on mobile.

            Since site layout is such a constant complaint (no matter what you guys do) why not put the html/CSS up on github and have a contest or vote?

            As far as content goes I personally would like more original content, more howto’s or introductions into subjects. For instance the article about the different types of glues was actually really useful to me. One of the comments took me to a youtube video about using baking soda and super glue to make a super super strong bond (I was able to secure this bolt I jam into my chair so it doesn’t tilt back SUPER SUPER well) Maybe it’s obvious to many people here but lots of newer folks have many gaps in our knowledge, we don’t know when to use one screw over another, or what type of glue to use. In many cases I don’t even know what to google. (and a reason I read hackaday, to expose me to things I didn’t know existed and terminology to I can research further) I’m not a big fan of the ‘controversial’ topics like 3d printing guns ( you know the ones that get like 200 comments of people arguing in 24 hours…) I just skip that stuff though.

            I realize the site is more for advanced stuff but I have noticed more tutorials and introductions as of late and if you’re going to do that go full blast.

            Some of the interesting discussions that arise from a comment may be worth posting. Some times I don’t know what the right answer is or want more opinions on it.

            Other content I would like is to know what the writers are doing. Why not a weekly or monthly post on what you guys are hacking on. (basically linked to

            I think the site should default to ‘blog’ but that’s just me. (autocomplete now goes to )


            – organize site content better
            – thread comments better and a better way to see new comments.
            – fix mobile site to be readable
            – more introduction/reference stuff (fasteners, tools, building techniques, tutorials, etc) or a wiki with key ‘you should know’ information

            All in all i’m happy with the content of the site. I like having more hacks (even if some of them are simple, there is always at least one ‘real’ hack) I think you guys are doing a good job over all and things have improved for the better since way back when everything was black in white.

          2. Holy crap, actual feedback from ProfessorK. That’s actually a pretty good list. Organizing the site…. we’re dealing with 10 years of technological debt, so that’s not going to be easy. It is, however, possible. Tutorials will be a thing once we get the hackaspace up and running. This is stuff that’s a five year plan, but it’s possible.

            All in all, great feedback.

      2. You clearly care to attack someone for a comment. I am grateful to Michael for sharing the build. My criticism was not an attack on character but on the general tendency for projects on HaD to feature a micro-controller when it may not be necessary. While your comment was a direct and dirty attack at me. If you bother to take the time to read it then you would see I made a suggestion about how I thought one could alter the build in a more simplistic fashion and for less cost but still accomplish automation. However I will admit that I didn’t know the peculiarities of pocky that lead to the original design choices. Michael was kind enough to enlighten me. I’d like to see things from your point of view but I can’t seem to get my head that far up my butt.

        1. Thanks for describing another way to build this thing! We used to know one way, now with your comment we have two!
          Folks, let’s not see posts like this as a criticism but instead as another way of solving the problem. The more, the better! There should be a purely mechanical one, and one with a Raspberry PI, and with an FPGA… come on!

      3. You can’t see a pencil dispenser that uses a stepper motor and not say it’s overkill. Even better, this time the article pointed it out so we didn’t have to drown in comments about it (there are few relative to others…) but was still political about it. I like.

    2. Actually in order to prevent pocky jams I had to utilize that “shaking” motion while the motors turns. I completely agree though that this could also be solved with a much cheaper and simpler analog circuit. One thing that I didn’t mention is that the motor/arduino/driver system is something that I already had to put together for something that’s actually work related. So it made sense just to reuse the same stuff.

      1. Don’t get my previous post wrong, I enjoy a good Rube Goldberg machine (not saying this is one), I did’t realize that the shaking motion was a necessity to prevent issue. By all means if you have something use it. Thank you for making the build and for sharing. You probably have noticed one has to have thick skin around the HaD comment’s section :)

    3. Sure maybe Arduino = Overkill, but Time = Money. If its quicker to slap something together with an Arduino you already have why would you waste time designing, sourcing, soldering, and making the analog equivalent? Sure it might costs a few bucks less overall, but you’ll end up spending more valuable time on a simple project. These things are for fun, not designed for mass production where the marginal manufacturing costs add up to be vastly larger than the engineering cost.

  5. your idea wouldnt cost 0.00 you need to think about the labor involved and the time invested, unless you are using slave labour which in that case how much is slave and what galaxy ur on. total power comp may not needed to be rated at watts but rememebr, “oppressed” arduino hackaday reader, thinks about how much calorie is needs to do work. obviously this is not what REALLY(Why is this in all-caps?) be featured on this website!

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.