You are required to read and agree to the below before accessing a full-text version of an article in the IDE article repository.

The full-text document you are about to access is subject to national and international copyright laws. In most cases (but not necessarily all) the consequence is that personal use is allowed given that the copyright owner is duly acknowledged and respected. All other use (typically) require an explicit permission (often in writing) by the copyright owner.

For the reports in this repository we specifically note that

  • the use of articles under IEEE copyright is governed by the IEEE copyright policy (available at
  • the use of articles under ACM copyright is governed by the ACM copyright policy (available at
  • technical reports and other articles issued by M‰lardalen University is free for personal use. For other use, the explicit consent of the authors is required
  • in other cases, please contact the copyright owner for detailed information

By accepting I agree to acknowledge and respect the rights of the copyright owner of the document I am about to access.

If you are in doubt, feel free to contact

Detection of Breakdowns using Historical Work Orders Data Analysis for Preventive Maintenance


Albert Bergström , Subaharan Kailayanathan , Saji Kamdod , Dmitrii Shabunin , Simon Monie´ , Martin Norrbom , Mobyen Uddin Ahmed

Publication Type:



Maintenance is an essential aspect of any industry working with machines. Breakdowns are expensive and can cause significant disruptions in the process line. Work orders provide essential information about breakdowns. However, this information can be challenging to comprehend and classify the root causation behind the breakdowns. In this project, work orders were analyzed using Natural Language Processing to find hidden patterns that show the root cause behind these breakdowns. Six different methods were used to analyze the data set: three unsupervised machine learning algorithms, and three supervised machine learning algorithms. For unsupervised learning, K-means, DBSCAN, and Hierarchical clustering were used for clustering. For supervised learning, Random Forest, Linear Regression, and Naive-Bayes were used for classification. After examining the results, the methods were unable to find any hidden patterns that could be related to the root cause of the breakdowns. It was concluded that changes to the data set should be made before there could be any substantial information gain from the free text since the description of the problem and solution has such a high amount of variations. Additionally, several suggestions were made related to improvements that could be made to improve the performance of a future analysis.


author = {Albert Bergstr{\"o}m and Subaharan Kailayanathan and Saji Kamdod and Dmitrii Shabunin and Simon Monie´ and Martin Norrbom and Mobyen Uddin Ahmed},
title = {Detection of Breakdowns using Historical Work Orders Data Analysis for Preventive Maintenance },
month = {January},
year = {2020},
url = {}