Identifier

etd-03312016-152816

Degree

Master of Science in Industrial Engineering (MSIE)

Department

Mechanical Engineering

Document Type

Thesis

Abstract

Inspection and rework are two important issues of quality control. In this research, an N-stage flowline is considered to make decisions on these two issues. When defective items are detected at the inspection station the items are either scrapped or reworked. A reworkable item may be repaired at the regular defect-creating workstation or at a dedicated off-line rework station. Two problems (end-of-line and multistage inspections) are considered here to deal with this situation. The end-of-line inspection (ELI) problem considers an inspection station located at the end of the line while the multistage inspection (MSI) problem deals with multiple in-line inspection stations that partition the flowline into multiple flexible lines. Models for unit cost of production are developed for both problems. The ELI problem is formulated for determining the best decision among alternative policies for dealing with defective items. For an MSI problem a unit cost function is developed for determining the number and locations of in-line inspection stations along with the alternative decisions on each type of defects. Both of the problems are formulated as fractional mixed-integer nonlinear programming (f-MINLP) to minimize the unit cost of production. After several transformations the f-MINLP becomes a mixed-integer linear programming (MILP) problem. A construction heuristic, coined as Inspection Station Assignment (ISA) heuristic is developed to determine a sub-optimal location of inspection and rework stations in order to achieve minimum unit cost of production. A hybrid of Ant-Colony Optimization-based metaheuristic (ACOR) and ISA is devised to efficiently solve large instances of MSI problems. Numerical examples are presented to show the solution procedure of ELI problems with branch and bound (B&B) method. Empirical studies on a production line with large number of workstations are presented to show the quality and efficiency of the solution processes involved in both ELI and MSI problems. Computational results present that the hybrid heuristic ISA+ACOR shows better performance in terms of solution quality and efficiency. These approaches are applicable to many discrete product manufacturing systems including garments industry.

Date

2016

Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Sarker, Bhaba

Share

COinS