Download this issue (.pdf)
back next

ASME Autonomous Material Sorter

In a time when ‘going green’ is in, many people make an extra effort to recycle waste products. But what happens to all that plastic, glass, steel and aluminum, and more importantly, how is it sorted? This is exactly what the University of Nebraska-Lincoln’s mechanical engineering students have been attempting to find out.

ABOVE: Kevin Watts, Kevin Riedl, Michael Head, and Brad Griffin assemble the autonomous material sorter.
Photo Courtesy: Brian Neilson

The 2010 American Society of Mechanical Engineers (ASME) Student Design Competition was held April 9-10 in St. Louis, Mo. Each team entered a version of an Autonomous Material Sorter (AMS) designed according to a tight set of guidelines. The AMS must be capable of sorting three objects each of four common recyclable materials: plastic, glass, steel and aluminum, with different dimensions assigned to each material. With a competitive time frame of five minutes, one may assume a large, complex machine is necessary for the job, but the sorting must take place in a mere 340x580x400 mm area, about the size of a microwave. At the judge’s signal, the materials are dumped in a hopper, a switch is toggled on, and the materials are sorted through the AMS into five labeled hoppers. The fifth hopper is labeled “unsorted.”

Armed with a VEX Robotics Design System, six mechanical engineering students have set out to design and build a light AMS that not only sorts properly, but quickly in order to achieve a maximum competition score. The first task was to set up a sorting model.

Working with this design model in mind, they turned to the recycling industry for ideas on how to transfer the materials from the hopper to the AMS. The first design was then tested using gravity to feed the materials from the hopper onto an inclined conveyor. The purpose of the conveyor was to not only distribute one item into the testing compartment, but also orient it for height testing.

However, isolating one object at a time proved a great challenge even without the additional complexity of orienting it. The design was modified many times by reconfiguring the hopper or reorienting the conveyor. Finally, faculty adviser Bill Dick suggested feeding the materials into the AMS parallel with the force of gravity instead of perpendicular as before.

To achieve this, a stopper was needed below the hopper. Two sets of rollers built of interlocked zip ties were best suited for the task. The front set of rollers would rotate toward each other downward, drawing samples into the AMS in a controlled manner, while the back set of rollers, in line with the first, would rotate upward, agitating the contents of the hopper. If the AMS did not sense any sample entering, it would reverse the role of the rollers, having the front set agitate and the back set feed.

This design proved successful and provided an additional surprise of properly orienting the plastic materials to be subjected to the height test. Then the team just needed to assemble all the feeding, testing, and sorting parts of the AMS and program it to run autonomously.

To test and sort the materials, the team closely followed the original sorting model. First, the metal detector can identify both the aluminum and steel samples. These two samples are further sorted by an inclined, magnetic conveyor, allowing the aluminum to slide off into its bin while transporting the steel up and into its own bin.

If the metal detector test comes back negative, photo sensors are employed to determine whether the height of the sample is blocking a laser beam. If so, the sample is sorted as plastic into its bin. Samples of a lesser height will be sorted by default as glass into its bin. This system then does not employ the unsorted bin. Since the AMS does not have a means to accurately identify a glass material, it may have been jeopardizing crucial points. Points were awarded on the following basis:

Score = (1000 x CS) – (1500 x IS) + (100 x US) – (3 x T) – (W/10) – (4000 x BG)

  • CS is the number of correctly sorted materials
  • IS is the number of incorrectly sorted materials
  • US is the number of unsorted materials
  • T is the time required to sort the waste products in seconds (300 maximum)
  • W is the total weight of the empty device in grams
  • BG is the number of glass containers broken

Team member Kevin Watts has worked extensively on the electrical components of the AMS. He designed and built the metal detector to fit in a standard circular electrical junction box. Since then he has put many hours into designing a filter for the audio frequencies emitted from the materials as they impact the base of the testing compartment. The filter ideally would be capable of positively identifying all four materials, not just distinguishing between plastic and glass as initially intended.

The entire team consists of seven members:

  • Michael Head - Chair
  • Brad Griffin - Vice Chair
  • Ryan McCormick - Team Adviser
  • Thomas Frederick - Chief Designer
  • Kevin Watts - Electrician
  • Kevin Riedl
  • Ben Stangl

For Watts and Griffin, designing the AMS has been more than competition; it is also their senior design project. The hands-on design is what attracted Stangl, and Riedl said he has enjoyed the “great teamwork and camaraderie.”

ABOVE: The partially finished autonomous material sorter
Photo Courtesy: Brian Neilson

Frederick’s expertise with the VEX Robotics Design System has also proved an invaluable asset to the team. As the AMS continues to take shape, Head indicated he is“extremely happy with the progress our team has made this year, and I was really excited to see us compete in April.”