Self-service systems potentially provide customers with continuous access to a service without the presence of on-site staff. As monitoring devices cannot perfectly detect every failure, the insufficient awareness of system states leads to either excessive service downtime or unnecessary maintenance costs. To assist service providers in determining the best time for maintenance, we propose an approach that incentivizes customers to report failures during service, thereby improving the overall awareness of system states and helping the service providers balance maintenance costs against fleet reliability. We develop a dynamic maintenance policy that uses time-varying thresholds of detected failures to trigger maintenance. We prove that the optimal threshold decreases over time but increases with higher customer incentives. By integrating imperfect automated monitoring with customer reports to maximize the fleet’s long-term profit, we capture system state transitions across periodic demand cycles through two stochastic processes. We reformulate the original mixed-integer nonlinear optimization problem into a tractable one and show that it can be decomposed into several subproblems, each with a single decision variable and a quasi-concave objective function. An iterative ternary search algorithm is designed to efficiently compute the exact optimal solution. A real-world case study on electric vehicle charging stations in Hong Kong demonstrates that the proposed maintenance policy achieves a 13% higher profit rate compared to ordinary maintenance policies without customer incentives, while the developed solution algorithms save over 50% of computation time compared to the Genetic Algorithm.

Professor Yao Cheng is an Assistant Professor in the Department of Data and Systems Engineering at The University of Hong Kong. She earned her PhD in Industrial Engineering from Rutgers, The State University of New Jersey. Leveraging tools from stochastic modeling and data analytics, Professor Cheng’s research primarily focuses on developing theoretical methodologies for reliability and resilience-based performance modeling and optimization, with applications in engineered systems across various sectors, including service systems and new energy systems. Her work has been published in and featured by leading journals in reliability and industrial engineering, such as IISE Transactions, Naval Research Logistics, Transportation Science, IEEE Transactions on Industrial Informatics, IEEE Transactions on Automation Science and Engineering, and Reliability Engineering and System Safety. Her research has been supported by funding from the Research Grants Council, the University Research Council, and industry partners.