Capstone team reduces plastic waste with the help of machine learning
(From left) Erian Perez-Gamboa, Malcolm West, Samantha Perry, Lauren Froberg, Shelby Janssen, David Sepulveda and Tej Scott comprise Team 25009, which is focusing on reducing 3D printer waste in the Engineering Design Center.
A recent study out of the University of Arizona Engineering Design Center found that more than a quarter of all 3D printer runs result from user error, wasting the plastic filament.
To address this issue, Team 25009 is fashioning a tool that can automatically adjust the 3D printer settings to match the needs of the design input. If a user selects the wrong plastic filament for a 3D printout, the tool can automatically override the incorrect settings and use the proper plastic source or it can stop the print process so the user can make a correction.
"It was a great chance to apply what I know in a totally different context," said team member Shelby Janssen. "The project has definitely met—and exceeded—my expectations. I’ve been able to build on my spectroscopy foundation and apply it to materials I haven’t worked with before. One of the most interesting and challenging aspects has been working with a low-cost spectral sensor."
The students used spectroscopy to identify plastic filaments like PLA, ASA, and PETG. They also had to develop a machine learning algorithm to accurately identify the materials.
"Like a lot of senior design teams, we've been developing our system at the Engineering Design Center," said the MSE senior. "It's a great resource, and it feels special to know we're developing a solution that can directly improve the space we've spent so much time in. Having access to the tools and resources at the EDC has made a huge difference, and it's fulfilling to think that our final product might improve the very space and tools that helped us build it."