Optimizing tool service life using tool vibration monitoring

Johannes Pan; Clemens Gutschi; Nikolaus Furian

Optimizing tool service life using tool vibration monitoring

Johannes Pan, Clemens Gutschi, Nikolaus Furian

Institute of Engineering and Business Informatics (3740)

Research output: Contribution to conference › Abstract › peer-review

Abstract

Numerous operations in manufacturing industry include machining activities, e.g. drilling, milling, turning, honing and more. This not only removes material from the work piece, but also wear the tool over time. Service life determines when the tool is replaced or comes to the revision. The service life of the tool is difficult to calculate precisely due to the different influences that the tool is exposed during work and thus results in a certain fluctuation range. Too early or too late intervention increases the setup time or is reflected in lack of part quality in case of too late intervention and is therefore costly. One of the keys to achieving maximum productivity in the use of tools is the individual determination of the condition of the tool over its service life for each individual. This study presents a method how to use the row data from vibration sensors placed on the tool holder. Statistical correlation analysis is used to extract features which show the correlation between the increasing wear of the tool and the produced product quality. Machine learning algorithms are used for automated diagnosis of tool service life to find the optimal time for use of tools.

Original language	English
Publication status	Published - Jun 2019
Event	30th European Conference on Operational Research: 30th European Conference on Operational Research - University College Dublin, Dublin, Ireland Duration: 23 Jun 2019 → 26 Jun 2019 https://www.euro2019dublin.com/

Conference

Conference	30th European Conference on Operational Research
Abbreviated title	EURO 2019
Country/Territory	Ireland
City	Dublin
Period	23/06/19 → 26/06/19
Internet address	https://www.euro2019dublin.com/

Keywords

Predictive maintenance
predictive analytics
random forest
ensemble prediction
machine learning
Probability of failure estimation
maintenance
Maintenance Management
Data analysis
data mining
Fast Fourier transforms
feature selection
feature engineering
Remaining useful life
life cycle assessment
condition monitoring

UN SDGs

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

Cite this

@conference{f6c4f21d6fcc4ef4a4810df9206f817f,

title = "Optimizing tool service life using tool vibration monitoring",

abstract = "Numerous operations in manufacturing industry include machining activities, e.g. drilling, milling, turning, honing and more. This not only removes material from the work piece, but also wear the tool over time. Service life determines when the tool is replaced or comes to the revision. The service life of the tool is difficult to calculate precisely due to the different influences that the tool is exposed during work and thus results in a certain fluctuation range. Too early or too late intervention increases the setup time or is reflected in lack of part quality in case of too late intervention and is therefore costly. One of the keys to achieving maximum productivity in the use of tools is the individual determination of the condition of the tool over its service life for each individual. This study presents a method how to use the row data from vibration sensors placed on the tool holder. Statistical correlation analysis is used to extract features which show the correlation between the increasing wear of the tool and the produced product quality. Machine learning algorithms are used for automated diagnosis of tool service life to find the optimal time for use of tools.",

keywords = "Predictive maintenance, predictive analytics, random forest, ensemble prediction, machine learning, Probability of failure estimation, maintenance, Maintenance Management, Data analysis, data mining, Fast Fourier transforms, feature selection, feature engineering, Remaining useful life, life cycle assessment, condition monitoring",

author = "Johannes Pan and Clemens Gutschi and Nikolaus Furian",

year = "2019",

month = jun,

language = "English",

note = "30th European Conference on Operational Research : 30th European Conference on Operational Research , EURO 2019 ; Conference date: 23-06-2019 Through 26-06-2019",

url = "https://www.euro2019dublin.com/",

}

TY - CONF

T1 - Optimizing tool service life using tool vibration monitoring

AU - Pan, Johannes

AU - Gutschi, Clemens

AU - Furian, Nikolaus

PY - 2019/6

Y1 - 2019/6

N2 - Numerous operations in manufacturing industry include machining activities, e.g. drilling, milling, turning, honing and more. This not only removes material from the work piece, but also wear the tool over time. Service life determines when the tool is replaced or comes to the revision. The service life of the tool is difficult to calculate precisely due to the different influences that the tool is exposed during work and thus results in a certain fluctuation range. Too early or too late intervention increases the setup time or is reflected in lack of part quality in case of too late intervention and is therefore costly. One of the keys to achieving maximum productivity in the use of tools is the individual determination of the condition of the tool over its service life for each individual. This study presents a method how to use the row data from vibration sensors placed on the tool holder. Statistical correlation analysis is used to extract features which show the correlation between the increasing wear of the tool and the produced product quality. Machine learning algorithms are used for automated diagnosis of tool service life to find the optimal time for use of tools.

AB - Numerous operations in manufacturing industry include machining activities, e.g. drilling, milling, turning, honing and more. This not only removes material from the work piece, but also wear the tool over time. Service life determines when the tool is replaced or comes to the revision. The service life of the tool is difficult to calculate precisely due to the different influences that the tool is exposed during work and thus results in a certain fluctuation range. Too early or too late intervention increases the setup time or is reflected in lack of part quality in case of too late intervention and is therefore costly. One of the keys to achieving maximum productivity in the use of tools is the individual determination of the condition of the tool over its service life for each individual. This study presents a method how to use the row data from vibration sensors placed on the tool holder. Statistical correlation analysis is used to extract features which show the correlation between the increasing wear of the tool and the produced product quality. Machine learning algorithms are used for automated diagnosis of tool service life to find the optimal time for use of tools.

KW - Predictive maintenance

KW - predictive analytics

KW - random forest

KW - ensemble prediction

KW - machine learning

KW - Probability of failure estimation

KW - maintenance

KW - Maintenance Management

KW - Data analysis

KW - data mining

KW - Fast Fourier transforms

KW - feature selection

KW - feature engineering

KW - Remaining useful life

KW - life cycle assessment

KW - condition monitoring

M3 - Abstract

T2 - 30th European Conference on Operational Research

Y2 - 23 June 2019 through 26 June 2019

ER -