Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data

Willy Dharmawan; Haitham Saleh; Nur Hamid; Sami El-Ferik; M. Yusha Firdaus; Choesnul Jaqin

doi:10.1109/IEEM63636.2025.11357747

Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data

Willy Dharmawan^*
, Haitham Saleh
, Nur Hamid
, Sami El-Ferik
, M. Yusha Firdaus
, Choesnul Jaqin

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

This study presents a machine learning-based approach for classifying noise types in automotive power seat DC motors using vibration data captured via a piezoelectric accelerometer. To construct a comprehensive feature representation, statistical features were extracted from both time and frequency domains, including RMS, kurtosis, spectral centroid, and Mel-frequency cepstral coefficients (MFCCs). The proposed framework evaluates ensemble learning classifiers - Random Forest, XGBoost, and AdaBoost - trained using stratified 5-fold cross-validation. Experimental results demonstrate that combining time- and frequency-domain features significantly enhances classification accuracy, with Histogram Gradient Boosting achieving the highest performance (95.58% accuracy, F1-score 0.9597). These findings highlight the effectiveness of multi-feature integration and ensemble methods for reliable, data-driven motor quality control in automotive manufacturing.

Original language	English
Title of host publication	IEEM 2025 - IEEE International Conference on Industrial Engineering and Engineering Management
Publisher	IEEE Computer Society
Pages	183-187
Number of pages	5
ISBN (Electronic)	9798331525217
DOIs	https://doi.org/10.1109/IEEM63636.2025.11357747
State	Published - 2025
Event	2025 IEEE International Conference on Industrial Engineering and Engineering Management, IEEM 2025 - Melbourne, Australia Duration: 7 Dec 2025 → 10 Dec 2025

Publication series

Name	IEEE International Conference on Industrial Engineering and Engineering Management
ISSN (Print)	2157-3611
ISSN (Electronic)	2157-362X

Conference

Conference	2025 IEEE International Conference on Industrial Engineering and Engineering Management, IEEM 2025
Country/Territory	Australia
City	Melbourne
Period	7/12/25 → 10/12/25

Bibliographical note

Publisher Copyright:
© 2025 IEEE.

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 9 Industry, Innovation, and Infrastructure

Keywords

Accelerometer data
Ensemble methods
Feature extraction
Machine learning
Motor noise classification

ASJC Scopus subject areas

Business, Management and Accounting (miscellaneous)
Safety, Risk, Reliability and Quality
Industrial and Manufacturing Engineering

Access to Document

10.1109/IEEM63636.2025.11357747

Cite this

Dharmawan, W., Saleh, H., Hamid, N., El-Ferik, S., Firdaus, M. Y., & Jaqin, C. (2025). Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data. In IEEM 2025 - IEEE International Conference on Industrial Engineering and Engineering Management (pp. 183-187). (IEEE International Conference on Industrial Engineering and Engineering Management). IEEE Computer Society. https://doi.org/10.1109/IEEM63636.2025.11357747

Dharmawan, Willy ; Saleh, Haitham ; Hamid, Nur et al. / Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data. IEEM 2025 - IEEE International Conference on Industrial Engineering and Engineering Management. IEEE Computer Society, 2025. pp. 183-187 (IEEE International Conference on Industrial Engineering and Engineering Management).

@inproceedings{ee31663b7d0d4e7b9abaeae8d9353b9a,

title = "Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data",

abstract = "This study presents a machine learning-based approach for classifying noise types in automotive power seat DC motors using vibration data captured via a piezoelectric accelerometer. To construct a comprehensive feature representation, statistical features were extracted from both time and frequency domains, including RMS, kurtosis, spectral centroid, and Mel-frequency cepstral coefficients (MFCCs). The proposed framework evaluates ensemble learning classifiers - Random Forest, XGBoost, and AdaBoost - trained using stratified 5-fold cross-validation. Experimental results demonstrate that combining time- and frequency-domain features significantly enhances classification accuracy, with Histogram Gradient Boosting achieving the highest performance (95.58\% accuracy, F1-score 0.9597). These findings highlight the effectiveness of multi-feature integration and ensemble methods for reliable, data-driven motor quality control in automotive manufacturing.",

keywords = "Accelerometer data, Ensemble methods, Feature extraction, Machine learning, Motor noise classification",

author = "Willy Dharmawan and Haitham Saleh and Nur Hamid and Sami El-Ferik and Firdaus, \{M. Yusha\} and Choesnul Jaqin",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 2025 IEEE International Conference on Industrial Engineering and Engineering Management, IEEM 2025 ; Conference date: 07-12-2025 Through 10-12-2025",

year = "2025",

doi = "10.1109/IEEM63636.2025.11357747",

language = "English",

series = "IEEE International Conference on Industrial Engineering and Engineering Management",

publisher = "IEEE Computer Society",

pages = "183--187",

booktitle = "IEEM 2025 - IEEE International Conference on Industrial Engineering and Engineering Management",

address = "United States",

}

Dharmawan, W , Saleh, H , Hamid, N , El-Ferik, S, Firdaus, MY & Jaqin, C 2025, Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data. in IEEM 2025 - IEEE International Conference on Industrial Engineering and Engineering Management. IEEE International Conference on Industrial Engineering and Engineering Management, IEEE Computer Society, pp. 183-187, 2025 IEEE International Conference on Industrial Engineering and Engineering Management, IEEM 2025, Melbourne, Australia, 7/12/25. https://doi.org/10.1109/IEEM63636.2025.11357747

Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data. / Dharmawan, Willy ; Saleh, Haitham ; Hamid, Nur et al.
IEEM 2025 - IEEE International Conference on Industrial Engineering and Engineering Management. IEEE Computer Society, 2025. p. 183-187 (IEEE International Conference on Industrial Engineering and Engineering Management).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data

AU - Dharmawan, Willy

AU - Saleh, Haitham

AU - Hamid, Nur

AU - El-Ferik, Sami

AU - Firdaus, M. Yusha

AU - Jaqin, Choesnul

PY - 2025

Y1 - 2025

N2 - This study presents a machine learning-based approach for classifying noise types in automotive power seat DC motors using vibration data captured via a piezoelectric accelerometer. To construct a comprehensive feature representation, statistical features were extracted from both time and frequency domains, including RMS, kurtosis, spectral centroid, and Mel-frequency cepstral coefficients (MFCCs). The proposed framework evaluates ensemble learning classifiers - Random Forest, XGBoost, and AdaBoost - trained using stratified 5-fold cross-validation. Experimental results demonstrate that combining time- and frequency-domain features significantly enhances classification accuracy, with Histogram Gradient Boosting achieving the highest performance (95.58% accuracy, F1-score 0.9597). These findings highlight the effectiveness of multi-feature integration and ensemble methods for reliable, data-driven motor quality control in automotive manufacturing.

AB - This study presents a machine learning-based approach for classifying noise types in automotive power seat DC motors using vibration data captured via a piezoelectric accelerometer. To construct a comprehensive feature representation, statistical features were extracted from both time and frequency domains, including RMS, kurtosis, spectral centroid, and Mel-frequency cepstral coefficients (MFCCs). The proposed framework evaluates ensemble learning classifiers - Random Forest, XGBoost, and AdaBoost - trained using stratified 5-fold cross-validation. Experimental results demonstrate that combining time- and frequency-domain features significantly enhances classification accuracy, with Histogram Gradient Boosting achieving the highest performance (95.58% accuracy, F1-score 0.9597). These findings highlight the effectiveness of multi-feature integration and ensemble methods for reliable, data-driven motor quality control in automotive manufacturing.

KW - Accelerometer data

KW - Ensemble methods

KW - Feature extraction

KW - Machine learning

KW - Motor noise classification

UR - https://www.scopus.com/pages/publications/105033959626

U2 - 10.1109/IEEM63636.2025.11357747

DO - 10.1109/IEEM63636.2025.11357747

M3 - Conference contribution

AN - SCOPUS:105033959626

T3 - IEEE International Conference on Industrial Engineering and Engineering Management

SP - 183

EP - 187

BT - IEEM 2025 - IEEE International Conference on Industrial Engineering and Engineering Management

PB - IEEE Computer Society

T2 - 2025 IEEE International Conference on Industrial Engineering and Engineering Management, IEEM 2025

Y2 - 7 December 2025 through 10 December 2025

ER -

Dharmawan W , Saleh H , Hamid N , El-Ferik S, Firdaus MY, Jaqin C. Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data. In IEEM 2025 - IEEE International Conference on Industrial Engineering and Engineering Management. IEEE Computer Society. 2025. p. 183-187. (IEEE International Conference on Industrial Engineering and Engineering Management). doi: 10.1109/IEEM63636.2025.11357747

Machine Learning-Based Noise Classification for Automotive Power Seat DC Motors Using Multi-Feature Representations from Accelerometer Data

Abstract

Publication series

Conference

Bibliographical note

UN SDGs

Keywords

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this