Tool condition monitoring for CNC milling operation via multisensor data fusion and machine learning

Norazam, Muhammad Azim Zufayri and Saleh, Tanveer (2026) Tool condition monitoring for CNC milling operation via multisensor data fusion and machine learning. In: 10th International Conference on Mechatronics Engineering (ICOM’26), 08-09 September 2026, IIUM.

PDF - Published Version Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

This paper presents an end-to-end tool condition monitoring (TCM) system for CNC milling that integrates multisensor data fusion, machine learning, and cloud-based remote visualization. Vibration signals at 1000 Hz and acoustic emission (AE) signals are processed through thirty two-second subwindows per minute, with features averaged to produce a single 44-element statistical and spectral feature vector per one-minute group, yielding 95 minute-level feature vectors across all seven machining runs. A Support Vector Machine (SVM) classifier with an RBF kernel achieves a group-based cross-validated accuracy of 82.9% ± 4.9% across three ISO 8688-2[10] wear states (New, Middle, and Worn), using a five-fold Stratified GroupKFold protocol that partitions data strictly by machining run to prevent temporal data leakage. The trained inference engine is deployed on a Flask server hosted on HuggingFace Spaces, paired with a Netlify web dashboard and HiveMQ MQTT telemetry streaming, completing an Industry 4.0 monitoring architecture accessible remotely without operator presence at the machine.

Item Type:	Proceeding Paper (Plenary Papers)
Uncontrolled Keywords:	CNC tool wear; SVM; acoustic emission; vibration; IoT; Industry 4.0; predictive maintenance; multisensor fusion
Subjects:	T Technology > T Technology (General)
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button):	Kulliyyah of Engineering > Department of Mechatronics Engineering
Depositing User:	Dr. Tanveer Saleh CEng MIMechE
Date Deposited:	30 Oct 2026 16:23
Last Update:	30 Oct 2026 16:23
Queue Number:	2026-10-Q5880
URI:	http://irep.iium.edu.my/id/eprint/132312
Indexed In:	Google Scholar

Actions (login required)

View Item