Location
Havener Center, Miner Lounge / Wiese Atrium, 9:30am-11:30am
Start Date
4-2-2026 9:30 AM
End Date
4-2-2026 11:30 AM
Presentation Date
April 2, 2026; 9:30am-11:30am
Description
Preventive Maintenance models have traditionally relied on a time-based maintenance model, which uses fixed statistical distribution to represent the time-to-failure (TTF). The requirement of data-driven models and automated decision-making systems have become essential for modern manufacturing systems. The use of fixed statistical distribution limits the ability of existing models in producing solutions in real-time. Our model overcomes these limitations by implementing an empirical distribution to represent TTF. The empirical distribution is generated using a neural network which eliminates noise from the raw maintenance log. Renewal Reward Theorem (RRT) is implemented to efficiently provide maintenance threshold in real time. The use of RRT is validated by comparing its performance with the discrete-event simulator. To support the decision-making process, we have implemented a human-in-the-loop framework. The framework enables users to update the threshold frequently and successfully incorporates online data-handling, optimization and decision-making in real-time, making it a robust model for preventive maintenance.
Biography
I am currently pursuing a Ph.D. in Systems Engineering at Missouri University of Science and Technology as a Kummer Innovation and Entrepreneurial Doctoral Fellow. I earned my bachelor's degree in industrial engineering from Tribhuvan University in Nepal in 2018.
My current research focuses on optimizing machine maintenance scheduling through preventive maintenance strategies. The proposed model integrates Renewal Reward Theory (RRT), and Artificial Intelligence within the context of Industry 4.0. The project shows results from integrating neural networks with RRT, and the validity of the model is confirmed by using Discrete Event Simulation (DES) in a controlled setting. These initial findings have strengthened our confidence in the project's potential for success. Future research will further explore and test the integration of Industry 4.0 technology using agent-based framework.
Meeting Name
2026 - Miners Solving for Tomorrow Research Conference
Department(s)
Engineering Management and Systems Engineering
Document Type
Poster
Document Version
Final Version
File Type
event
Language(s)
English
Rights
© 2026 The Authors, All rights reserved
Preventive Maintenance Scheduling Using Artificial Intelligence and Decision Support Agent
Havener Center, Miner Lounge / Wiese Atrium, 9:30am-11:30am
Preventive Maintenance models have traditionally relied on a time-based maintenance model, which uses fixed statistical distribution to represent the time-to-failure (TTF). The requirement of data-driven models and automated decision-making systems have become essential for modern manufacturing systems. The use of fixed statistical distribution limits the ability of existing models in producing solutions in real-time. Our model overcomes these limitations by implementing an empirical distribution to represent TTF. The empirical distribution is generated using a neural network which eliminates noise from the raw maintenance log. Renewal Reward Theorem (RRT) is implemented to efficiently provide maintenance threshold in real time. The use of RRT is validated by comparing its performance with the discrete-event simulator. To support the decision-making process, we have implemented a human-in-the-loop framework. The framework enables users to update the threshold frequently and successfully incorporates online data-handling, optimization and decision-making in real-time, making it a robust model for preventive maintenance.
Comments
Advisor: Abhijit Gosavi, gosavia@mst.edu