TY - JOUR
T1 - Engine replacement scheduling optimization using Data Mining
AU - Farizal,
AU - Joelian, Albert
N1 - Funding Information:
This work has been supported partially by PITTA 2019 Grant funded by DRPM Universitas Indonesia under contract No: NKB-0732/UN2.R3.1/HKP.05.00/2019.
Publisher Copyright:
© 2020 IOP Publishing Ltd. All rights reserved.
PY - 2020/5/28
Y1 - 2020/5/28
N2 - Engine overhaul activity in heavy duty equipment takes long shutdown duration, while unscheduled replacement is impacted on process delay, increasing man power cost, and production loss. One main cause of these problems is the scheduling performed just based on mechanics' intuition and experience. On the other side, condition monitoring data are available in a large number. Reliable data processing methods are needed to disclose hidden information from the data. For the purpose, this research used three data mining methods on condition monitoring data and external factors of heavy equipment engine to get an optimized engine replacement scheduling. Clustering method was used to classify condition monitoring data, association rule was used to analyze the interrelationship between variables and time series analysis was used to predict the value of condition monitoring. The result showed that data mining methods can be used to perform scheduling optimization. Unscheduled replacement engine or engine failed in service was reduced from 31% to 13%.
AB - Engine overhaul activity in heavy duty equipment takes long shutdown duration, while unscheduled replacement is impacted on process delay, increasing man power cost, and production loss. One main cause of these problems is the scheduling performed just based on mechanics' intuition and experience. On the other side, condition monitoring data are available in a large number. Reliable data processing methods are needed to disclose hidden information from the data. For the purpose, this research used three data mining methods on condition monitoring data and external factors of heavy equipment engine to get an optimized engine replacement scheduling. Clustering method was used to classify condition monitoring data, association rule was used to analyze the interrelationship between variables and time series analysis was used to predict the value of condition monitoring. The result showed that data mining methods can be used to perform scheduling optimization. Unscheduled replacement engine or engine failed in service was reduced from 31% to 13%.
UR - http://www.scopus.com/inward/record.url?scp=85086320429&partnerID=8YFLogxK
U2 - 10.1088/1742-6596/1500/1/012111
DO - 10.1088/1742-6596/1500/1/012111
M3 - Conference article
AN - SCOPUS:85086320429
VL - 1500
JO - Journal of Physics: Conference Series
JF - Journal of Physics: Conference Series
SN - 1742-6588
IS - 1
M1 - 012111
T2 - 3rd Forum in Research, Science, and Technology International Conference, FIRST 2019
Y2 - 9 October 2019 through 10 October 2019
ER -