Optimizing risk management strategies of reverse osmosis for pharmaceutical purified water systems through enhanced integration of machine learning and quantitative risk assessment /
Nur Danisha Amalia Binti Zainuddin, , ,
Optimizing risk management strategies of reverse osmosis for pharmaceutical purified water systems through enhanced integration of machine learning and quantitative risk assessment / Nur Danisha Amalia Binti Zainuddin - xi, 56 pages : illustrations ;
Faculty of Chemical and Process Engineering Technology
Final Year Report (Bachelor of Manufacturing Engineering Technology (Pharmaceuticals) with Hons ) -- Universiti Malaysia Pahang Al-Sultan Abdullah - 2025
Include bibliographical reference
In the pharmaceutical industry, ensuring consistent product quality requires effective quality risk management (QRM) due to the high risks associated with processes and products. A significant challenge arises in purified water systems, particularly in reverse osmosis (RO) units, where membrane fouling affects production reliability. This study explores how big data analytics and machine learning (ML) can improve risk assessment for RO systems used in water purification for pharmaceuticals. This research utilized the Friday 13th (Fr 13) framework with @RISK software for quantitative risk analysis and WEKA software for machine learning simulations. The study began by identifying a suitable mathematical model for the RO process. Risk assessments were performed using the Fr 13 framework with @RISK software, simulating 10,000 possible risk scenarios. Following this, a systematic methodology was employed to validate the p-values from @RISK in WEKA, training the random forest algorithm by separating 70% of the data for training and 30% for testing. The results revealed a minimal 0.43% difference in risk failure predictions between WEKA and @RISK, highlighting the reliability of WEKA’s machine learning capabilities. The findings emphasize that machine learning can enhance the efficiency and accuracy of risk assessments, offering a more adaptable and resource-efficient approach compared to traditional methods. By integrating these advanced tools, this research contributes to achieving UN Sustainable Development Goal (SDG) 6 (Clean Water and Sanitation), ensuring high-quality water for pharmaceutical manufacturing. To conclude, WEKA's machine learning algorithms offer efficient risk assessment, saving time and resources compared to manual and conventional approaches.
THE0010744 (Local)
Faculty of Chemical and Process Engineering Technology--Dissertations
Universities and colleges--Dissertations
Theses--Dissertations
Optimizing risk management strategies of reverse osmosis for pharmaceutical purified water systems through enhanced integration of machine learning and quantitative risk assessment / Nur Danisha Amalia Binti Zainuddin - xi, 56 pages : illustrations ;
Faculty of Chemical and Process Engineering Technology
Final Year Report (Bachelor of Manufacturing Engineering Technology (Pharmaceuticals) with Hons ) -- Universiti Malaysia Pahang Al-Sultan Abdullah - 2025
Include bibliographical reference
In the pharmaceutical industry, ensuring consistent product quality requires effective quality risk management (QRM) due to the high risks associated with processes and products. A significant challenge arises in purified water systems, particularly in reverse osmosis (RO) units, where membrane fouling affects production reliability. This study explores how big data analytics and machine learning (ML) can improve risk assessment for RO systems used in water purification for pharmaceuticals. This research utilized the Friday 13th (Fr 13) framework with @RISK software for quantitative risk analysis and WEKA software for machine learning simulations. The study began by identifying a suitable mathematical model for the RO process. Risk assessments were performed using the Fr 13 framework with @RISK software, simulating 10,000 possible risk scenarios. Following this, a systematic methodology was employed to validate the p-values from @RISK in WEKA, training the random forest algorithm by separating 70% of the data for training and 30% for testing. The results revealed a minimal 0.43% difference in risk failure predictions between WEKA and @RISK, highlighting the reliability of WEKA’s machine learning capabilities. The findings emphasize that machine learning can enhance the efficiency and accuracy of risk assessments, offering a more adaptable and resource-efficient approach compared to traditional methods. By integrating these advanced tools, this research contributes to achieving UN Sustainable Development Goal (SDG) 6 (Clean Water and Sanitation), ensuring high-quality water for pharmaceutical manufacturing. To conclude, WEKA's machine learning algorithms offer efficient risk assessment, saving time and resources compared to manual and conventional approaches.
THE0010744 (Local)
Faculty of Chemical and Process Engineering Technology--Dissertations
Universities and colleges--Dissertations
Theses--Dissertations