Accurate Characterization of Mixed Plastic Waste Using Machine Learning and Fast Infrared Spectroscopy

Stas Zinchik; Shengli Jiang; Søren Friis; Fei Long; Lasse Høgstedt; Victor M. Zavala; Ezra Bar-Ziv

doi:10.1021/acssuschemeng.1c04281

Accurate Characterization of Mixed Plastic Waste Using Machine Learning and Fast Infrared Spectroscopy

Stas Zinchik, Shengli Jiang, Søren Friis, Fei Long, Lasse Høgstedt, Victor M. Zavala, Ezra Bar-Ziv

Research output: Contribution to journal › Article › peer-review

27 Scopus citations

Abstract

We present a combination of convolutional neural network (CNN) framework and fast MIR (mid-infrared spectroscopy) for classifying different types of dark plastic materials that are commonly found in mixed plastic waste (MPW) streams. Dark plastic materials present challenges in fast identification because of the low signal-to-noise ratio. The proposed CNN architecture (which we call PlasticNet) can reach an overall classification accuracy of 100% and can identify the constituent materials in a multiplastic blend with 100% accuracy. The fast MIR system can collect spectral data at a rate up to 400 Hz, and the CNN model can reach prediction speeds of 8200 Hz. Therefore, this method provides an avenue to be able to characterize MPW in a real-time high-throughput manner.

Original language	English
Pages (from-to)	14143-14151
Number of pages	9
Journal	ACS Sustainable Chemistry and Engineering
Volume	9
Issue number	42
DOIs	https://doi.org/10.1021/acssuschemeng.1c04281
State	Published - 25 Oct 2021
Externally published	Yes

Keywords

Classification
IR spectra
Machine learning
Plastic waste
Real-time

ASJC Scopus subject areas

General Chemistry
Environmental Chemistry
General Chemical Engineering
Renewable Energy, Sustainability and the Environment

Access to Document

10.1021/acssuschemeng.1c04281

Cite this

@article{6647603b3af542028bfb677e1fbec324,

title = "Accurate Characterization of Mixed Plastic Waste Using Machine Learning and Fast Infrared Spectroscopy",

abstract = "We present a combination of convolutional neural network (CNN) framework and fast MIR (mid-infrared spectroscopy) for classifying different types of dark plastic materials that are commonly found in mixed plastic waste (MPW) streams. Dark plastic materials present challenges in fast identification because of the low signal-to-noise ratio. The proposed CNN architecture (which we call PlasticNet) can reach an overall classification accuracy of 100% and can identify the constituent materials in a multiplastic blend with 100% accuracy. The fast MIR system can collect spectral data at a rate up to 400 Hz, and the CNN model can reach prediction speeds of 8200 Hz. Therefore, this method provides an avenue to be able to characterize MPW in a real-time high-throughput manner.",

keywords = "Classification, IR spectra, Machine learning, Plastic waste, Real-time",

author = "Stas Zinchik and Shengli Jiang and S{\o}ren Friis and Fei Long and Lasse H{\o}gstedt and Zavala, {Victor M.} and Ezra Bar-Ziv",

note = "Publisher Copyright: {\textcopyright} 2021 American Chemical Society.",

year = "2021",

month = oct,

day = "25",

doi = "10.1021/acssuschemeng.1c04281",

language = "English",

volume = "9",

pages = "14143--14151",

journal = "ACS Sustainable Chemistry and Engineering",

issn = "2168-0485",

publisher = "American Chemical Society",

number = "42",

}

TY - JOUR

T1 - Accurate Characterization of Mixed Plastic Waste Using Machine Learning and Fast Infrared Spectroscopy

AU - Zinchik, Stas

AU - Jiang, Shengli

AU - Friis, Søren

AU - Long, Fei

AU - Høgstedt, Lasse

AU - Zavala, Victor M.

AU - Bar-Ziv, Ezra

PY - 2021/10/25

Y1 - 2021/10/25

N2 - We present a combination of convolutional neural network (CNN) framework and fast MIR (mid-infrared spectroscopy) for classifying different types of dark plastic materials that are commonly found in mixed plastic waste (MPW) streams. Dark plastic materials present challenges in fast identification because of the low signal-to-noise ratio. The proposed CNN architecture (which we call PlasticNet) can reach an overall classification accuracy of 100% and can identify the constituent materials in a multiplastic blend with 100% accuracy. The fast MIR system can collect spectral data at a rate up to 400 Hz, and the CNN model can reach prediction speeds of 8200 Hz. Therefore, this method provides an avenue to be able to characterize MPW in a real-time high-throughput manner.

AB - We present a combination of convolutional neural network (CNN) framework and fast MIR (mid-infrared spectroscopy) for classifying different types of dark plastic materials that are commonly found in mixed plastic waste (MPW) streams. Dark plastic materials present challenges in fast identification because of the low signal-to-noise ratio. The proposed CNN architecture (which we call PlasticNet) can reach an overall classification accuracy of 100% and can identify the constituent materials in a multiplastic blend with 100% accuracy. The fast MIR system can collect spectral data at a rate up to 400 Hz, and the CNN model can reach prediction speeds of 8200 Hz. Therefore, this method provides an avenue to be able to characterize MPW in a real-time high-throughput manner.

KW - Classification

KW - IR spectra

KW - Machine learning

KW - Plastic waste

KW - Real-time

UR - http://www.scopus.com/inward/record.url?scp=85118192259&partnerID=8YFLogxK

U2 - 10.1021/acssuschemeng.1c04281

DO - 10.1021/acssuschemeng.1c04281

M3 - Article

AN - SCOPUS:85118192259

SN - 2168-0485

VL - 9

SP - 14143

EP - 14151

JO - ACS Sustainable Chemistry and Engineering

JF - ACS Sustainable Chemistry and Engineering

IS - 42

ER -

Accurate Characterization of Mixed Plastic Waste Using Machine Learning and Fast Infrared Spectroscopy

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this