- 200 views
- 208 downloads
Using Machine Learning and Keyword Analysis to Analyze Incident Reports and Reduce Risk in Oil Sands Operations
-
- Author / Creator
- Kurian, Daniel G
-
Many companies maintain large databases of incident reports. Incidents that have severe consequences are analyzed in detail to prevent recurrence, while minor incidents are typically stored without any further evaluation. Especially with common incidents and those with lesser consequences, details that are necessary to understand the cause of the incident might be missing. Incidents that occur in the oil and gas industry can be reported more accurately and analyzed to provide value to companies maintaining databases, and to prevent and mitigate risks. Such information can be used to lower costs and improve safety culture.
The initial objective of this study was to create a risk matrix system for collectively analyzing incident reports, commensurate across companies, for increased reliability in reporting and enhanced analytical power across an industry. A supervised machine learning approach was applied in conjunction with this risk matrix to analyze incident reports and provide a risk score.
During this research project, 15,000 incident reports, including both process and occupational-type incidents, were analyzed from three oil sand companies across Alberta. The results were classified by incident type (determined by industry experts) and consequence type (using the risk matrix). Furthermore, potential and actual risk scores were evaluated for every incident using the risk matrix. This analysis built the foundation for a system to identify trends and leading indicators, and to design prevention and mitigation strategies across the entire industry.
The goals of this researched evolved to include the application of artificial intelligence and machine learning to create a digitalized system for efficiently reporting incidents that can be used to generate a risk matrix, trend report, prevention and mitigation strategies, and leading indicator identification for every incident report that is inputted.
Implementing this system was accomplished by utilizing a combination of supervised machine learning and keyword analysis. During this research project, the 15,000 incident reports were analyzed to build a customized library of keywords. These keywords were assigned to a list of statements that were generated using a company’s safety guidelines, standard operating procedures, and asset management systems. The basic structure for generating outputs was demonstrated using a large incident database provided by collaborators of the project and some sample inputs. Three case studies of incident reports were also processed and presented using the proposed methodology, delivering practical outputs that could be used by workers and companies to improve safety and increase hazard awareness.
-
- Graduation date
- Spring 2020
-
- Type of Item
- Thesis
-
- Degree
- Master of Science
-
- License
- Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.