Play Live Radio
Next Up:
0:00
0:00
0:00 0:00
Available On Air Stations

Recycling, trash or compost? An Austin sixth grader is using AI to sort his school's garbage.

A boy in a striped t-shirt holds a lemon in front of a laptop camera that is sitting on a wooden stand.
Patricia Lim
/
KUT News
Cyril Summerfield, a student at Marshall Middle School, demonstrates the waste detector he designed to help sort trash.

An Austin ISD sixth grader is developing a computer program to determine whether trash belongs in the recycling, compost or landfill bins.

Cyril Summerfield had been volunteering to help custodians sort waste in the cafeteria after lunch at General Marshall Middle School. The school, like the district as a whole, is committed to significantly reducing its waste by 2040.

"Cyril spent a couple of lunch periods there going, 'You know what? I think I can figure out something easier than having kids stand up here during their lunchtime,’” said Stacey Smith, a librarian who is the Green Team coordinator at Marshall. "'They could have more time to talk to their friends.'"

Cyril said sorting trash by hand left something to be desired.

“It wasn’t the most entertaining job in the world," he said, "so I designed this thing which uses machine learning and [artificial intelligence] to decide whether it’s trash, recycling or compost and then tell you that.”

Cyril developed the computer program using the website Machine Learning for Kids and Scratch, a free programming language from the Massachusetts Institute of Technology. He’s still in the process of training it to recognize what is trash, recycling or compost.

“It gets it wrong like an eighth of the time, so I have to go, 'No, that’s recycling,’ and then it will think about it for a second," he said. "And then I do that a few more times and it will get it right."

ML Waste Detector GT Showcase Video.mp4

Cyril demonstrated how the program works by holding up different items to a webcam. He held up a lemon and the program correctly identified it as “compost.” When the program identifies waste incorrectly, he corrects it, so the program gets it right next time.

Cyril has a good sense of humor about his creation, which on occasion has identified him as “trash" when he shows up on camera.

“And when it says ‘compost,’ I’m like, 'You’re right,” he said.

Machine learning burns aside, Cyril said it’s important to him that students understand where waste belongs to protect the environment.

“We really have to make sure kids are putting the right things into the right bins so we don’t have tons of waste going into the landfill,” he said. “In fact, when compost goes into the landfill it releases methane, which is a worse greenhouse gas than what is coming out of your car.”

A boy wearing a striped t-shirt holds a web cam that he is trying to place inside of a wicker basket.
Patricia Lim
/
KUT News
Cyril works on a portable prototype he has created to help sort waste in his school's cafeteria.

Smith said she was "delighted" to see the students want to help solve the trash problem.

Cyril said he thinks all kids think about climate change and what it will mean for their future. In the immediate future, Cyril plans to keep perfecting his waste-sorting program.

“I plan to do a ton of work on this over the summer," he said, "and then maybe have it as a product by the start of [the] next school year."

Becky Fogel is the education reporter at KUT. Got a tip? Email her at rfogel@kut.org. Follow her on Twitter @beckyfogel.
Related Content