Lower Wavelength Near Infrared (NIR) Spectroscopy for Affordability; A New Technology for Waste Management Scalability: Near Infrared (NIR) Spectroscopy

My STS project draws conclusions from the work I did for my computer engineering
Capstone project. For this project, my team designed a near-infrared (NIR) spectrometer, using
lower wavelength LED emitters and photodiode detectors than the traditional industrial-grade
systems on the market. The goal was to create a proof of concept for the future scalability of NIR
spectrometers, with respect to waste management systems specifically. Part of my research also
includes the benefits of using such technology within waste management industries.

The technical portion of my Thesis investigated the testing of the actual spectrometry
work involved in building my Capstone device. The system, named the PlastiClass Express, was
designed to classify three different polymers of plastic: HDPE, PET, and PP by measuring the
reflectance of the plastic sample. This was done by pulsing LEDs at 3 distinct wavelengths of
near-infrared light while taking measurements with a photodiode detector. These measurements
were processed through an embedded code algorithm and the classified polymer gets displayed
on an LCD screen. All of this was built into a 3D-printed enclosure which also contained a
special enclosure for the optics hardware. Although it was expected that the system would be
able to classify the three polymers by calculating reflectance ratios between the distinct
wavelengths, we found that both PET and PP were incorrectly identified, unlike HDPE. As a
result, certain scaling adaptations were made to our embedded code algorithm to compensate for
this, and the PlastiClass Express succeeded in classifying the three polymers to a certain degree.
Thus, we were able to conclude that the PlastiClass Express was a functional prototype for
performing near-infrared spectrometry using lower wavelengths, enabling future spectrometry
development to consider saving money on hardware costs.
The STS research conducted investigated some of the societal benefits near-infrared
spectrometry has already proven when implemented into waste management systems. This
technology comes into play as these systems need to sort waste, and there are various ways to do
so. Near-infrared spectrometry outperforms another powerful technology, image-based deep
learning, which is both less accurate and less efficient in terms of processing speed. The cost
structure of near-infrared spectrometry is advantageous too, as quality control costs can be cut by
up to 90% with respect to “people time and error.” These are just a few examples of the value
that near-infrared spectrometry brings to waste management, as more attention ought to be
brought to such a technology that is making a positive impact on the environment.

Working on the Capstone and technical portions of my research first helped me develop
insights regarding the relationship between technology and environmental awareness. Cutting
costs in one area can lead to improvements elsewhere in waste management, which need to be
developed in the continuing fight to improve the environment. By scoping out the various
implications of near-infrared spectrometry, I’m now able to draw from my technical work and
understand the types of systems where this technology would be integrated, and I would
encourage those within the waste management industry to continue and improve upon this work.
Finally, I would like to acknowledge my Capstone team: Jack Chandler and Eric Powell,
as well as my Capstone and STS professors, Harry J. Powell and Richard Jacques respectively.

School of Engineering and Applied Science
Bachelor of Science in Computer Engineering
Technical Advisor: Harry J. Powell
STS Advisor: Richard Jacques
Technical Team Members: Jack Chandler, Eric Powell