It won't separate plain from peanut, but a robot programmed by a Ball State University research team can sort out red, green, yellow, orange, blue and brown M&Ms.

The project won second place in a nationwide contest sponsored by Parallax, a robotics company based in Rocklin, California.

Chris Hileman, a senior majoring in manufacturing engineering technology, teamed up with Rick Johnson, an assistant professor in the Department of Industry and Technology, to program the robot.

Johnson purchased a kit that included a color sensor, microprocessor and servo motor.

Hileman programmed the device to collect an M&M from a tube, determine its color and move it into another tube with M&Ms of the same color.

"Some of the other entrants were teams of two or three people and for Chris to be able to manage this complex project on his own was very impressive," said Parallax spokesman Eric Wood.

Working with such a tasty test subject was difficult.

"Chris kept accusing me of sabotage because every time an M&M would pop out of the machine we'd eat it," Johnson said.

The contest used M&Ms because the candy is easy to find and work with.
 
"Parallax was looking for something with multiple colors," Johnson said. "An M&M has a fairly hard shell and it's easy to come by. The flat shape is important. Skittles won't go through the machine because they're too round."

Hileman said orange M&Ms caused the fewest errors.

"The machine likes orange because the red value measures high, so it's easy to distinguish," he said. "Orange sticks out like a sore thumb."

Hileman used the contest as an independent study program as part of his efforts to make himself more marketable to future employers.

"From a resume perspective, it gives credence to an employer that Chris is capable of competing," Johnson said. "He can take minimal instruction from a blank slate and create a viable project."

The team won $150 in cash and $100 worth of hardware from Parallax.

For more information, contact Johnson at (765) 285-5656 or rjohnson@bsu.edu. For video of the robot, contact Peter Osborne at (765) 285-1563.