Science Fair Project Sorts Recyclables

This crew of high schoolers built a sorting robot for the Smart Young Mindz challenge. We got pretty excited when hearing that it sorts plastic by its recycling code, but unfortunately this isn’t quite what it’s made out to be. The device uses an RFID code on each product to figure out where it goes. Their thinking is that at some point every product sold will have an embedded tag in it. For now this will not revolutionize the recycling industry, but the build is still impressive. We’re sure they learned a ton from all of the mechanical engineering that went into the project.

You can see the three laundry baskets that serve as the sorting bins. The white box above the bin on the right is the hopper in which a plastic container is placed. The box can then revolve around a central axis to position itself over the correct basket. The floor of the box is then retracted, dropping the refuse in the bin. Check out the video after the break for the satisfying cry of the servo motors at work.

We like seeing recycling robots, but so far most of what we’ve seen are aluminum can crushers.

19 thoughts on “Science Fair Project Sorts Recyclables

    1. Africa is not the backwards place that his comment presumes/implies that it is, and I would thank you to not be condescending. Although this project is not a perfect solution, the thought process it represents is no more or less impressive than if it was that of a kid with similar knowledge in any other country.

  1. There are mechanical standards that make this possible now with no RF. It’s already used in automated recycle stations in Europe etc..

    For things that don’t match specs it’d take chemistry or photons or RFID. I hope RFID is never used, RF is already over-used for the sake of profit margins…

  2. Why couldn’t you simply scan the UPC and do a lookup of what product codes come in what container types? Then you wouldn’t need RFID and it would work immediately (although you may need to build a database for types if it’s product that isn’t known).

    1. UPC barcodes are only used in the US, The rest of the world use EAN.
      It is easy to look up the barcode data to what material the packaging is made from, If you have access to GS1 the data is there. The problem you will run into is if the barcode is not readable.

      The way RVM machines sort an item is via an internal database of barcode / weight / camera profile/color (

      1. Excellent post.

        I wasn’t thinking when I said UPC… but as you point out, surely there’s another barcode or other id that could be read with either a scanner or even a high speed webcam with some software analysis.

        Even if the database wasn’t present, you could build it yourself. Once a type is identified, it wouldn’t have to be identified again. The workload for the human would get less and less in theory.

        If it’s not readable, I would either toss it back into the hopper to see if it can get picked up on another pass… or sort it into an “unidentified” bin that gets manually sorted. At least then you’re lessening the workload for the human.

  3. Recycling is such a scam. You do all the work, big companies make all the money. If recycling was oh so valuable, then all of the mining companies would be buying up landfills instead of digging thru hard rock.

  4. To people who ask why recycle: Grade school chemistry and environmental science teaches what carbon and ecology is, you must of not been there… American am-i-right?

    Barcodes, RFID, Mechanical-spec matching are all heavily flawed methodologies. Either invest in unattended and accelerated decomposition supplemented with human analyses, or stick to recycling…

    1. Yeah, there are some really, really ignorant people out there. What’s the name of that “law” that the least competent are the most confident in their own knowledge and ability?

      1. Yeah it’s me who doesn’t know what carbon and ecology is…

        In my country stupid people only know two languages and are still doing linear algebra by ten. Nice pseudo-philosophy you got there though.

  5. All the technology to fully and automatically sort waste has existed for over 20 years, it just hasn’t all been applied together in a single processing plant.

    Start with smashing and shredding the waste into small chunks. Next step run it down conveyors with eddy current generators in the end rollers. Ferrous metals get pulled around the bend then drop as they move out of the magnetic field. Nonferrous metals get repelled and drop into another chute. Non-metallic waste drops straight down.

    With vibrators on the belts to separate stuff that sticks together and running the separated materials through a few stages of eddy current generators the result will be quite clean collections of ferrous and nonferrous metals, ready for sale to smelters.

    Liquids drained from the shredder could include all kinds of recyclable chemicals, motor oil and who knows what. It could be sold to refineries or filtered and used as fuel for other processes in the recycling plant.

    The next material to separate out would be glass. Glass is heavier than most plastics and other non-metallic waste. An easy way to get it out would be with a “fluidized bed”. That’s basically a slanted shaker table with a lot of small holes in the bottom through which air is blown, just enough to float everything lighter than glass. The glass chunks shake off below a splitter blade down one chute while everything else goes over the splitter.

    Want to color sort the glass? That’s easy. Run it down a mesh conveyor, spread into a thin layer. A camera looks for brown or green and other non-clear chunks then the system triggers an air jet under the belt to blow the chunk upward and towards one edge of the belt. Send brown left and green right. Clear chunks near the edges get puffed towards the center. Put multiple stages along a single belt and at the end there’s clear chunks pouring off the center and colors on the edges. Set the collection chutes to get a bit of clear into the colors because clear glass doesn’t foul up colored glass like a few chunks of colored glass can ruin a batch of clear.

    Now it’s down to the hard stuff, that yucky mess of organics and plastics that hangs together. Some of it floats, some of it doesn’t. Start with giving it a bath. The water used for the first stage wouldn’t have to be clean. Skim the floaty stuff off the top and the sinking stuff can be scraped out of the bottom with a chain driven scraper, up a slope at one end. (Such have long been used in sewage treatment plants.)

    Some of the water could be obtained from the liquids collected from the shredder at the beginning. A centrifuge would be a good way to separate liquids that aren’t water soluble. Remember the gizmos used to help clean up the Deepwater Horizon spill? Exactly the same principle as a rotary cream separator.

    Plastics that don’t float can be sorted using a bubble tank. Different size air bubbles stick to different types of plastic and lift them in water. That’d be another multi-stage process to sort out different types.

    One thing to note about plastics, the most desirable to recycle are the thermoplastics because they can be remelted and remolded, though the physical and mechanical plastics always change, so the recycled plastic cannot be used in the same way. It has to be mixed with “virgin” plastic or the product has to be designed to work with the altered properties.

    Thermoset or chemically catalyzed plastics cannot be melted again. They’re only usable as ground up powder or slivers to be used as a filler or burned as fuel.

    At the end of all the automated sorting processes, municipal solid waste (that’s what the stuff you toss in a trash can is called) becomes high quality, clean metal, glass, reusable plastics, chemical refinery feedstock and industrial process fuel.

    Whatever is burnable and not not going to be reused can be burned to provide heat for any drying steps, especially after sorting and cleaning processes that use water.

    That brings in yet another process, useful for handling sewage. Mix the sewage and non-recyclable burnables with just enough of a clay (such as Bentonite) to hold it together into pellets. Feed the pellets into the bottom of a high temperature cyclonic furnace, fired by burnable liquids collected and separated from the shredder. Run the pellet feed chute through the furnace exhaust duct to dry them before they hit the burner. As they burn out, the clay hardens and the pellets swirl upwards and are collected. The heat they give off as they cool would be useful for drying stages in other processes. The air intake for the furnace stage should be pulled from the waste dump room, shredder and anywhere else in the plant where odor would be an issue.

    What to do with the hardened, porous pellets? They’d make excellent aggregate for concrete and asphalt because the cement and tar would penetrate the pellets and thus bond to them far better than to rock and gravel.

    Put together all the available technology and a “push button” automatic recycling plant could be built that takes in solid waste and raw sewage then outputs clean water and clean metals and other materials, ready to recycle. The only thing it wouldn’t produce is a lot of expensive jobs to run the facility.

    1. all your methods fail against stuff with epoxies, corrosives, and mechanical oversizes.

      When you factor in the chemistry, mechanical engineering, and logistics, you’ll see it’ll take heavy chemical processing or humans.

  6. That’s very good idea to create standard database of UPC codes in South Africa. And also your video are very helpful. I think it’s a better Science fair project which is very informative to everyone so keep updating your blog readers by publishing interesting articles on your blog.

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.