AUTONOMOUS METHOD FOR DETECTING CUTTING TOOL AND MACHINE TOOL ANOMALIES IN MACHINING

Lead Research Organisation: University of Sheffield
Department Name: Automatic Control and Systems Eng

Abstract

In advanced manufacturing, there exists a rising demand for both high productivity and producing high-performance components with tighter tolerances. In order to meet these requirements, monitoring cutting tool conditions and machine tool health is needed to improve dimensional accuracy of workpiece, extend the cutting tool life, minimise machine tool down time and eliminate scrap and re-work costs.

Traditionally, monitoring cutting tool conditions and machine tool health is carried out by operators who perform a manual inspection, which often causes unnecessary stoppages of machine tools and, as a result, costs incurred from lost productivity. However, without a timely inspection of both cutter status and machine tool working conditions, cutter wear or breakage and machine tool malfunction can take place during machining causing significant damage to workpieces. Some researchers have estimated that the amount of machine tool downtime due to these problems is around 6.8% while others put the figure closer to 20%. Therefore, manufacturing costs can be significantly higher than necessary when either cutters are changed before the end of their useful life or after cutter wear and breakage or machine tool malfunction have caused damage to workpieces. Consequently, a real time and automatic inspection of cutting tool status and machine tool health conditions is needed to profoundly address these problems.

This project aims to propose a fundamental solution to the challenges faced by current technologies and develop innovative techniques that can autonomously detect cutting tool and machine tool anomalies in machining for advanced manufacturing. This innovative solution will be based on a novel approach known as sensor data modelling and model frequency analysis, which is uniquely developed by the PI's team at Sheffield and has recently found applications in the condition monitoring and fault diagnosis of a wide range of engineering systems and structures.

The project will involve a close multi-disciplinary collaboration of ACSE academics, AMRC engineers, and industrial partners. The novel project idea and this unique research collaboration are expected to fundamentally resolve many challenges and produce urgently needed diagnostic technologies for autonomously detecting cutting tool and machine tool anomalies in machining for advanced manufacturing industry in UK.

Planned Impact

The proposed project will develop an innovative sensor data modelling and model frequency analysis-based approach, which would create a paradigmatic shift to the current way of monitoring of cutting tool condition and machine tool health, realising autonomously detecting cutting tool and machine tool anomalies in machining for advanced manufacturing. Thanks to more robust to noises and use of physically meaningful features, the new approach can more effectively reveal cutting tool physical characteristic changes and machine tool spindle conditions. This would fundamentally address challenges faced by existing visual aid or direct sensor signal analysis-based monitoring technologies. The potential impacts of the research outcomes are as follows:

-The project will open a new research area of direct industrial relevance for system and control scientists as well as provide innovative monitoring technologies for researchers working on cutting tool and machine tool operation and maintenance. The project will study and apply a novel nonlinear model frequency analysis based structural system feature extraction. This is also of significance for the application of machine learning to condition monitoring and fault diagnosis of a wide range of complicated engineering systems and structures where feature extraction is a challenging but necessary and important step.

-The developed approach can be extended to address process monitoring related research challenges in many other industrial processes including, e.g., safety critical systems such as robotic arms used in nuclear power plant maintenance and quality critical manufacturing such as drilling in aircraft manufacturing.

-The autonomous diagnostic system developed in the project is expected to result in radical improvements in both cost efficiency and product quality in advanced manufacturing, allowing work to move back to the UK from low-wage economies, strengthening UK supply chains, and boosting the development of the UK economy.

-The market opportunity for the developed diagnostic system is significant as metal machining represents 10% and remains the most important of all manufacturing processes in the UK. The adoption of the new approach based technologies will produce direct benefits to many industrial sectors in the UK, particularly, the UK aerospace industry which has recently seen sales of £35.9bn with a work force of over 111,000 (Source: UK ADS 2018) .

These cover a wide range of impact areas including academic, economic, industrial, and related social impacts.

Publications

10 25 50
 
Description The aim of the project is to propose a fundamental solution to the challenges faced by current technologies and develop innovative techniques that can autonomously detect cutting tool and machine tool anomalies in machining for advanced manufacturing.

So far, the achievements of the project can be summarised as follows :

- We have derived a novel methodology for industrial process monitoring and validated the effectiveness and performance of the methodology using existing experimental data collected by AMRC engineers at Sheffield. Three quality journal papers about the new methodology have already been published.

- We have, based on the novel industrial process monitoring methodology, developed a range of innovative implementation techniques for the detection of cutting tool and machine tool anomalies in advanced manufacturing processes. We have successfully completed two rounds of industrial scale experimental studies at ARMC to carry out real time monitoring of cutting tool conditions in machining processes: one for a milling process designed according to generic needs in advanced manufacturing industry and the other for a milling process designed for manufacturing products of interests to our industrial partners. We have showcased the industrial scale experimental study results to our industrial partners. They have all shown great interests in our achievements and provided valuable advice regarding how to develop these techniques to higher TRL levels.

- We have completed all spindle test rig related laboratory works at AMRC as planned in the grant proposal. This has provided us with sufficient real time data that will enable us to extend the application of our novel industrial process monitoring methodology to a variety of industrial areas.

The plan for the next stage of the project is to engage with industrial partners in advanced manufacturing and relevant sectors via industrial application-oriented case studies to disseminate research outcomes and realise expected impact.
Exploitation Route The final outcomes of this funding would be the basis of a range of autonomous diagnostic systems that can be widely used for the detection of cutting tool and machine tool anomalies in advanced manufacturing.
Sectors Aerospace

Defence and Marine

Digital/Communication/Information Technologies (including Software)

Manufacturing

including Industrial Biotechology

 
Description A keynote speech at the International Conference on Knowledge-Based and Intelligent Information & Engineering Systems (KES-23) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact This was a key note talk on: "Machine Learning Methods for Sensor Data Fusion and Autonomous Systems".
The Conference on Knowledge-Based and Intelligent Information & Engineering Systems (KES-23) provides leading edge events on various aspects of Sustainable Technology including sustainable buildings, smart energy, sustainable design and manufacturing. It has
a community consisting of several thousand research scientists, academics, engineers and practitioners who participate in KES activities.

My talk stimulated a discussion about autonomous systems, deep learning methods and evaluation of their resilience.
Year(s) Of Engagement Activity 2023
URL http://www.kesinternational.org/
 
Description How to Increase Autonomy with Machine Learning Methods, a keynote talk for the 10th International Conference on Signal Processing and Integrated Networks (SPIN 2023) 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Postgraduate students
Results and Impact Abstract of the talk
There is a fast development of different machine learning methods - for object classification, tracking, action recognition and other tasks with multiple types of data - from images and videos to time series data. Autonomous image and video analytics faces a number of challenges due to the huge volumes of data that sensors provide, the changeable environmental conditions and other factors. However, it is important to know when the methods work well and when they are not reliable, e.g. how much could we trust the obtained results? How could we characterise trust is a related question. How could we quantify the impact of uncertainties on the developed solutions? This talk discussed current trends in the area of machine learning and show results for image and video analytics for autonomous systems.
Automated detection and behaviour analysis is another important area which necessitates unsupervised learning algorithms. Recent results for automated video analytics were presented with Dirichlet process models, deep learning and other methods. Their pros and cons were discussed.
Year(s) Of Engagement Activity 2022
URL https://www.amity.edu/spin2023/
 
Description RAS 2024 | 7th Annual IEEE UK and Ireland Robotics and Automation Society Chapter Conference 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact I chaired a session of Autonomy, Trust and verification.
Also I gave a talk as part of this session, giving an overview of our research related to these topics.
Year(s) Of Engagement Activity 2024
URL https://www.sheffield.ac.uk/sheffieldrobotics/7th-ieee-uk-ireland-ras-conference-ras-2024
 
Description Talk for UK and EU industrial engagements 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact To share the research outcomes with the UK and EU industrial members of AMRC to realise expected impact and benefit the UK advanced manufacturing industry.
Year(s) Of Engagement Activity 2023
 
Description Talk for international industrial engagements 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Industry/Business
Results and Impact To share the research outcomes with an international industrial member of AMRC to realise expected impact and benefit the UK advanced manufacturing industry.
Year(s) Of Engagement Activity 2023
 
Description Talks at University of British Columbia and University of Alberta in Canada for research collaborations 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Study participants or study members
Results and Impact To share the research outcomes with Canadian scientists working on advanced manufacturing and related areas for international collaborations.
Year(s) Of Engagement Activity 2023
 
Description Towards Trustworthy Machine Learning Methods 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Results and Impact This is an invited talk as part of the Terrorism Risk Assessment, Modelling and Mitigation Seminar Series (TRAMMSS) at Cranfield University, 11 November 2022
Year(s) Of Engagement Activity 2022
URL https://www.cranfield.ac.uk/events/events-2022/towards-trustworthy-machine-learning-methods