GOALI: Stochastic Optimization Framework for Energy-Smart Re/Manufacturing Systems

点击翻译按钮获取中文摘要

This Grant Opportunities for Academic Liaison with Industry (GOALI) award will contribute to the national welfare by developing models to support the efficient integration of manufacturing and remanufacturing production lines. Remanufacturing is important to sustainable production by extending product life and reducing the environmental impact of manufacturing. Uncertain customer demand, along with highly variable product returns in both quantity and quality, have proved challenging to manufacturers in planning to allocate production capacity between new and returned products. The project, a collaboration between University of Louisville, Northeastern University, and IBM Corporation, will consider production scheduling and inventory levels, energy impact, uncertainty in demand, returns quantity, and returns quality to produce a production plan that is scalabile to industry-scale problems. The researched modeling approach is expected to inform the way such hybrid systems are designed, operated, and sustained, and will promote awareness of manufacturing-related e-waste considerations. The project will benefit US manufacturing by enabling the development of best practices for production-inventory management and recommendations for energy consumption and minimization of the energy footprint.This research will develop a novel three-stage stochastic optimization model that integrates tactical (production and energy) and operational (inventory) decisions under a single integrated framework. The third-stage operational decisions reflect three levels of uncertainty (demand, returns quantity, and returns quality). The second-stage, NP hard server-to-bank allocation problems (in the second stage) is addressed through a dual bin-packing model approach. The overall solution approach employs a scenario-based decomposition framework. A high-fidelity simulation model for the overall system will allow benchmarking of real-world strategies against solutions generated by the new approach. The industrial partner will pilot an implementation of the most promising policy from the benchmarking exercise, which will enable translation of the findings and fine-tuning of the approach. The project contributes to the training of next generation engineers via computational tools (e.g., optimization, virtual reality and simulation) and case studies to complement in-class instruction.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.