What is it about?
This research addresses energy consumption in milling machines, which are essential manufacturing tools that consume substantial electricity and contribute significantly to industrial carbon emissions. The study develops a systematic methodology to monitor and analyze energy usage patterns in these machines using K-means clustering, an unsupervised machine learning algorithm. Data were collected over three months from a PAMA Speedram 2000 milling machine, capturing approximately 204,000 power consumption measurements. The clustering algorithm successfully identified three distinct operational states: shutdown, idle/finishing operations, and roughing operations. Analysis revealed that auxiliary systems, particularly the chip conveyor and treatment unit, consume significant energy even during low-productivity periods. The study demonstrates that implementing duty-cycle modifications for the chip conveyor based on operational states can reduce total energy consumption by 46-53% during finishing operations and 11-18% during roughing operations. This approach provides manufacturers with data-driven insights to optimize energy usage without compromising productivity, offering a practical pathway toward more sustainable manufacturing practices through improved control of auxiliary subsystems.
Featured Image
Photo by Kevin Ku on Unsplash
Why is it important?
This research provides manufacturing industries with practical, immediately implementable strategies to reduce energy consumption and environmental impact. Given that manufacturing accounts for approximately 23% of global energy use, optimizing machine tool operations represents a significant opportunity for emissions reduction. The methodology presented enables manufacturers to identify energy waste in existing equipment without requiring costly machine replacements, addressing the constraint that machine tools typically operate for 20 years or more. The subsystem-level analysis framework can be applied across diverse manufacturing facilities to support the transition toward greener production strategies while maintaining economic competitiveness through reduced operational costs.
Perspectives
This study advances the application of unsupervised machine learning to industrial energy management by demonstrating how clustering algorithms can autonomously identify operational states from raw power consumption data. The approach bridges theoretical data science methods with practical manufacturing optimization, offering actionable insights for energy efficiency improvements in discrete, event-driven production environments. Future research should extend this methodology to multi-machine systems, incorporate real-time monitoring capabilities, and evaluate seasonal variability effects on energy profiles to enhance generalizability across different manufacturing contexts.
Sunil Maurya
Universita degli Studi di Firenze
Read the Original
This page is a summary of: K-Means Clustering Algorithm–Based Energy Profiling of Milling Machines: Status Based Optimization of Energy, Smart and Sustainable Manufacturing Systems, October 2025, ASTM International,
DOI: 10.1520/ssms20240025.
You can read the full text:
Contributors
The following have contributed to this page
