鶹ý

 

Graduate Student Research Seminar Day ‑ March 23, 2022

You are cordially invited to the Graduate Student Research Seminar of the Department of Industrial Engineering.

Date: Wednesday, March 23, 2022
Time: 1:00 - 3:20 PM
Venue:This is a hybrid event.
In-person gathering: Room MA310, Sexton Campus

Schedule:

1:00-1:25 PMHamzea Al-Jabouri
Optimal selective maintenance decisions for integrated imperfect large-scale selective maintenance and repairpersons assignment problem using a column generation approach.

1:25-1:50 PMLuana Almeida
Trends and gaps in the literature of road network repair and restoration in the context of disaster response operations

1:50-2:15 PMMd Hasan Ali
A simulation-optimization approach for robust job shop scheduling with condition-based maintenance and random breakdowns

2:15-2:30 PM Break

2:30-2:55 PMIbrahim Oguntola
Optimal design of a multimodal and sustainable supply chain network with shipment consolidation

2:55-3:20 PMBrian Song
Determining centralized inventory SKUs in a space and constrained hospital environment using MILP

Abstracts:

Optimal selective maintenance decisions for integrated imperfect large-scale selective maintenance and repairpersons assignment problem using a column generation approach
Hamzea Al-Jabouri, PhD Candidate

This talk presents a novel approach based on column-generation to maximize the reliability of a large-scale series-parallel system in the joint selective maintenance and repairperson assignment problem (JSM–RAP). The system performs consecutive missions separated by scheduled finite-duration breaks and the components are imperfectly maintained during the breaks. Current selective maintenance (SM) models either assume that only one repair channel is available or deal with a limited-scale of SM model with multiple repairing channels. The proposed approach decomposes the problem into a restricted master problem that is solved to update the Lagrangian multipliers and a set of sub-problems that are solved to generate maintenance patterns. Two methods are proposed to solve the mixed-integer nonlinear sup-problems: a piecewise-linear approximation, and a reformulation into a conic exponential program. Numerical experiments show the added value of the proposed approach in terms of computation time and the overall quality of solutions. Results also highlight the benefit of jointly carrying out the selection of the components to be maintained, the maintenance level to be performed and the assignment of the maintenance tasks to repairpersons.

Trends and gaps in the literature of road network repair and restoration in the context of disaster response operations
Luana Almeida, PhD Candidate

Natural and man-made disasters can disrupt road networks. The post-disaster period is usually divided into short and long-term response phases. In the short-term phase, blocked roads can jeopardize emergency operations such as search-and-rescue, evacuation of victims, and the distribution of emergency supplies. The detritus may pose an environmental challenge in the long-term, as the rubble must be properly collected, disposed, and recycled. Optimization models designed to schedule repair and restoration activities and routing repair crews are helpful to give insights to decision-makers on how to improve emergency response plans. This article presents a literature review on short and long-term network models designed for roads repair and restoration. This review identifies the trends in publications, solution approaches, frequency of locations and type of disaster in the case studies, and the size of the networks. In addition, a qualitative analysis of the models’ characteristics, such as the type of network, uncertainties, and interdependencies, is presented. This review shows that the field is growing rapidly through the work of separate research groups. A discussion is provided to indicate future research directions. Among these, further discussions on the disaster management phases considered in the models would be beneficial to further advance the research area. In addition, future research should focus on the combination of roads repair and restoration with other emergency activities such as relief distribution.

A simulation-optimization approach for robust job shop scheduling with condition-based maintenance and random breakdowns
Md Hasan Ali, MASc Student

The integration of scheduling and maintenance activities is important in shop floor decision-making. In reality, machines might be unavailable due to the maintenance activities in the scheduling horizon. To solve a Job Shop Scheduling Problem (JSSP), a simulation-optimization approach is proposed under both planned and stochastic machine unavailability in this study. Two maintenance policies are considered: Condition-Based Maintenance (CBM) and Corrective Maintenance (CM). The objective function of real makespan is first approximated using several surrogate functions and independently optimized using Genetic Algorithm (GA). The approximated and optimized schedule is then evaluated through simulation considering stochastic degradation of machines, random breakdowns, and uncertain CBM and CM duration. A weighted average of the expected makespan and its 90th percentile is used as the evaluation criterion to ensure schedule robustness, and the best schedule is added to an elite list to initiate the next iteration. To prevent a premature conversion, a stopping rule motivated by Simulated Annealing (SA) is employed in the outer loop of the proposed approach. The proposed approach showed good performance with a maximum average improvement of 10.87% for 15 jobs and 15 machines and a minimum average improvement of 5.80% for 50 jobs and 15 machines over the initial solution with an average run time of 2.5 hours and 8.3 hours respectively. Numerical experimentation demonstrated that the proposed approach can effectively generate high-quality schedules.

Optimal design of a multimodal and sustainable supply chain network with shipment consolidation
Ibrahim Oguntola, MASc. Student

Shipment Consolidation (SCL) is a factor that can improve optimality in the network design of sustainable multimodal logistics supply chains (SCs). SCL is a logistics strategy whereby multiple shipments are combined and dispatched as a single unit to receivers usually in the same market region. The SCL facilities such as make-bulk or break-bulk terminals between SC entities allow the network to take advantage of economies of scale. We formulate a mixed integer linear programming model of a multi-echelon SC network that considers carbon and water footprints, multimodal transportation and SCL simultaneously. Multiple SC network configurations considering different SCL strategies such as shipper-performed, or carrier-performed consolidation and direct delivery are comparatively analysed. We then investigate ideas such as the impact of the location of the SCL facilities relative to SC entities (e.g., supplier, manufacturer, or 3PL) on the total cost and environmental impact of an SC network. The influence of company policy on the relative importance of sustainability to economic viability on the best choice of SCL configuration is also explored. Results from multiple experiments using realistically generated data are then discussed

Determining centralized inventory SKUs in a space and constrained hospital environment using MILP
Brian Song, MASc Student

This research determines centralized inventory SKUs in a space and cost constrained hospital environment using Mixed Integer Linear Programming (MILP). The case study takes place in IWK Children’s Hospital located in Halifax where the demands of consumable medical supplies significantly increased since the outbreak of COVID 19. With increased demands and number of SKUs they must control, IWK wishes to investigate how their centralized inventory can be more efficient and effective at fulfilling demands of clinical critical items and prevent stockouts. Therefore, this research proposes a generic MILP model with Knapsack principles to determine centralized inventory SKUs based on space, cost and essentiality of items. The model is developed generic to be applicable to any healthcare sectors but be able to effectively reflect the uniqueness of each hospital by controlling parameters. The result suggests the list of centralized inventory SKUs with appropriate inventory policies: order quantity, reorder points and safety stocks. Furthermore, characterization of results is proposed with recommendation on setting inventory policies for new SKUs.

Contact Person:
Prof. Dr. Floris Goerlandt
email: floris.goerlandt@dal.ca