Bottleneck adjacent matching heuristics for scheduling a re-entrant flow shop with dominant machine problem

• Main Author: Sh Ahmad, Sh Salleh
• Format: Thesis
• Language: English; English; English
• Published: 2009
• Subjects: TS Manufactures; TS155-194 Production management. Operations management
• Online Access: http://eprints.uthm.edu.my/7448/1/24p%20SH%20SALLEH%20SH%20AHMAD.pdf; http://eprints.uthm.edu.my/7448/2/SH%20SALLEH%20SH%20AHMAD%20COPYRIGHT%20DECLARATION.pdf; http://eprints.uthm.edu.my/7448/3/SH%20SALLEH%20SH%20AHMAD%20WATERMARK.pdf; http://eprints.uthm.edu.my/7448/
• Institution: Universiti Tun Hussein Onn Malaysia

The re-entrant flow shop environment has become prominent in the manufacturing industries and has recently attracted researchers attention. Typical examples of re-entrant flow shops are the printed circuit board manufacturing and furniture painting processes where components or processed parts enter some specific machines more than once. Similar with other manufacturing environment, identifying appropriate scheduling methodologies to ensure high output rate is very much desirable. The problem explored and investigated in this research is a special type of scheduling problem found in a re-entrant flow shop where two of its processes have high tendency of exhibiting bottleneck characteristics. The scheduling problem resembles a four machine permutation re-entrant flow shop with the routing of M1,M2,M3,M4,M3,M4 where Ml and M4 have high tendency of being the dominant machines. The main objective of this research is to take advantage of the bottleneck characteristics at the re-entrant flow shop and use it to develop heuristics that can be used to solve its scheduling problems. There are four major concentrations in this research. First, basic mathematical properties or conditions that explain the behaviour of the bottleneck processes were developed to give an insight and clearer understanding of the re-entrant flow shop with dominant machines. Second, four new and effective scheduling procedures which were called BAM1 (Bottleneck Adjacent Matching 1), BAM2, BAM3 and BAM4 heuristics were developed. Third, bottleneck approach was utilised in the study and the analysis using Visual Basic macro programming indicated that this method produced good results. Fourth, the Bottleneck Scheduling Performance (BSP) indexes introduced in the BAM heuristics procedure could be used to ascertain that some specific generated job arrangements are the optimum schedule. As a general conclusion, this research has achieved the objectives to develop bottleneck-based makespan algorithms and heuristics applicable for re-entrant flow shop environment. The experimental results demonstrated that the BAM heuristics generated good performances within specific P1 (first process) bottleneck dominance level and when the number of jobs increases. Within the medium to large-sized problems, BAM2 is the best at weak PI dominance level whereas BAM4 is the best at strong P1 dominance level.