This research explores how artificial intelligence and data-driven methods can be applied to CNC machine tools to monitor equipment condition and support predictive, rather than purely reactive or fixed-schedule, maintenance.
The Problem
CNC machine tools are prone to unplanned downtime when wear or faults go undetected until they cause a failure, while purely time-based maintenance schedules can waste effort servicing equipment that doesn't yet need it.
Objectives
Approach
The work combines Python-based data analysis and machine learning with industrial condition-monitoring data, building models that learn what healthy and degrading equipment states look like.
Development
Current development is focused on data collection and preparation, feature analysis, and early model experimentation.
Testing
Models are being evaluated against recorded equipment condition data to assess how reliably they distinguish normal operation from developing faults.
This research is in active development, with a prototype approach being refined rather than a finished production system.
What’s Next
Future work includes expanding the underlying dataset, refining model accuracy, and exploring integration with live CNC condition monitoring.
Hotspot Management System
A hotspot management system featuring voucher authentication, internet packages, active-session monitoring and shared bandwidth allocation across multiple access points.
Engineering Systems / Technology Education
Exploring Python-driven CNC automation and G-code workflows to strengthen practical, hands-on engineering learning.
Energy Systems & Thermodynamics Research
Thermodynamic optimization of energy systems, including working-fluid selection and heat-recovery configuration for industrial applications.
If a project you're working on overlaps with any of this, I'd like to hear about it.