“Hey, You Left The Peanut Out Of My Peanut M&Ms!”

Candy-sorting robots are in plentiful supplies on these pages, and with good reason — they’re a great test of the complete suite of hacker tools, from electronics to machine vision to mechatronics. So we see lots of sorters for Skittles, jelly beans, and occasionally even Reese’s Pieces, but it always seems that the M&M sorters are the most popular.

This M&M sorter has a twist, though — it finds the elusive and coveted peanutless candies lurking in most bags of Peanut M&Ms. To be honest, we’d never run into this manufacturing defect before; being chiefly devoted to the plain old original M&Ms, perhaps our sample size has just been too small. Regardless, [Harrison McIntyre] knows they’re there and wants them all to himself, hence his impressive build.

To detect the squib confections, he built a tiny 3D-scanner from a line laser, a turntable, and a Raspberry Pi camera. After scanning the surface to yields its volume, a servo sweeps the candy onto a scale, allowing the density to be calculated. Peanut-free candies will be somewhat denser than their leguminous counterparts, allowing another servo to move the candy to the proper exit chute. The video below shows you all the details, and more than you ever wanted to know about the population statistics of Peanut M&Ms.

We think this is pretty slick, and a nice departure from the sorters that primarily rely on color to sort candies. Of course, we still love those too — take your pick of quick and easy, compact and sleek, or a model of industrial design.

20 thoughts on ““Hey, You Left The Peanut Out Of My Peanut M&Ms!”

  1. You put a bunch of them on an elevated flat surface, like a desktop. Then you take a wooden stick and you sweep them all over the edge. As they reach the edge, they will all have the same velocity. They should all fall in nice parabolic arcs. The lighter ones will land closer. A judiciously arranged pair of collection boxes should instantly sort out the lighter ones, no? Just a thought. It’s your basic mass spectrometer. If this works (easy to test), you could easily automate it much more simply. Use an inclined board, doesn’t have to be very wide. Drop the particles off at the top, one or a few at a time, and let gravity accelerate them to the bottom, where you put a curve in the board like a water slide, so they fly off flat. Same pair of boxes. No matter their weight, all reach the same “terminal” velocity, and Newton handles the rest. Only flaw I see in all this is if the weight difference between the two types is not enough to create a separation distance that is larger than the noise in the system.

    1. Why stop at a peanut M&M mass spectrometer? Why not make a peanut M&M supercollider and smash them together to analyze their contents? Sure, you’ve destroyed the M&Ms, but you’d know exactly what they’re made of!

    2. You’re counting on their cross-section slowing them down in the X direction before they hit the ground? They’d better be dropping pretty far!

      How about a vacuum cleaner tube, calibrated for just the right flow that separates them? Or a fan pointed up a downward sloping table?

    3. I am wrong. This won’t work. On re-look, the parabolas are dependent only on the x-velocity and not mass. Ignoring friction or air resistance, my idea will not work. The distance out from the edge is x0 + x0*v0 + (1/2)gt^2, which has no “m”s in it. (Nor “n”s). The v0 is the bottom-of-the-track exit v, which (ignoring friction) will be identical for all. Sorry… Headstone reads “It seemed like a good idea at the time.” ;)

Leave a Reply to Daniel Dunn Cancel 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.