Search for: All records

Computer Server Industry is characterized by extensive test processes to ensure high quality and reliability of the servers. Computer Server Industry production systems utilize Configure-To-Order (CTO), also known as fabrication/fulfillment, strategy which provides an effective balance between demand and supply by synchronizing the flow of materials, equipment, and labor throughout the production process. In the fabrication stage, components or sub-assemblies are produced, tested, and assembled based on a projected production plan. They are then kept in stock until an actual order is received from a customer. In the fulfillment stage, final products are assembled according to actual customer orders. Assignment of products to test cells during the fulfillment stage can be a challenging task due to high quality requirement and limited resources. Current practices tend to assign products to test cells based on a specific criterion such as on-time shipment or maximum test cell occupancy, which can result in higher levels of energy consumption or delayed orders. This paper introduces a Deep Reinforcement Learning (DRL) approach to effectively assign servers to test cells considering a multi-objective reward function that combines multiple criteria. A proposed simulation model serves as the environment with which the DRL agent interacts, learning a policy that develops a test schedule for the products. The proposed approach is tested with a case study from a high-end server manufacturing environment. Sensitivity analysis is conducted to analyze the impact of the different values of the system’s variables on its performance. 

Trauma presents a prominent health problem worldwide. However, trauma centers are often clustered in urban areas and sparsely located in rural areas. The geographic maldistribution of trauma centers leads to system-related mistriage errors. While some local governments oer subsidy to incentivize the affiliated hospital group to redesign the trauma care network, the approach is ad hoc. To address this issue, we propose a bilevel integer programming model to investigate the subsidized trauma care network redesign problem, which considers the government as the leader and the hospital group as the follower. To solve the resultant problem efficiently, we propose a branching idea to exclude additional infeasible solutions and suboptimal solutions, in turn speeding up the branch-and-bound algorithm. In a case study, we redesign a trauma care network in the midwestern area of the U.S. based on closed-form approximate functions of system-related mistriage errors. The results show that the optimal network redesign redistributes the network by slightly reducing the number of trauma centers to relieve the crowded trauma care resource, and achieves an overall improvement of about 11% over the original network. 

Trauma injuries continue to be the leading cause of mortality and morbidity among US citizens aged 44 years and under. Government agencies are often in charge of designing an effective trauma network in their region to provide prompt and definitive care to their citizens. This process is, however, largely manual, experience-based and often leads to a suboptimal network in terms of patient safety. To support effective decision making, we propose a Nested Trauma Network Design Problem (NTNDP), which can be characterized as a nested multi-level, multi-customer, multi-transportation, multi-criteria, capacitated model with the bi-objective of maximizing the weighted sum of equity and effectiveness in patient safety. We use mistriages (system-related under- and over-triages) as surrogates for patient safety. To add realism, we include intermediate trauma centers that are set up in many states in the US to serve as feeder centers to major trauma centers to improve patient safety and three criteria to mimic EMS’s on-scene decisions. We propose a ‘3-phase’ solution approach that first solves a relaxed version of the model, then solves a Constraint Satisfaction Problem, and then a modified version of the original optimization problem (if needed), all using a commercial solver. Our findings suggest that solutions are sensitive to (i) the proportion of assignments attributed to various destination determination criteria, (ii) distribution of trauma patients, and (iii) relative emphasis on equity vs. effectiveness. We also illustrate the use of our approach using real data from a midwestern US state; results show over 30% performance improvement in the objective value.