Modeling and analysis of multiobjective lot splitting for N‐product M‐machine flowshop lines

Abstract

Lot splitting is a new approach for improving productivity by dividing production lots into sublots. This approach enables accelerating production flow, reducing lead-time and increasing the utilization of organization resources. Most of the lot splitting models in the literature have addressed a single objective problem, usually the makespan or flowtime objectives. Simultaneous minimization of these two objectives has rarely been addressed in the literature despite of its high relevancy to most industrial environments. This work aims at solving a multiobjective lot splitting problem for multiple products in a flowshop environment. Tight mixed-integer linear programming (MILP) formulations for minimizing the makespan and flowtime are presented. Then, the MinMax solution, which takes both objectives into consideration, is defined and suggested as an alternative objective. By solving the MILP model, it was found that minimizing one objective results in an average loss of about 15% in the other objective. The MinMax solution, on the other hand, results in an average loss of 4.6% from the furthest objective and 2.5% from the closest objective.

A review of lot streaming in a flow shop environment with makespan criteria

“…For the problem of m/N/C/CV a heuristic and the use of GA for sequencing the products and for determining the number of sublots were proposed (Kumar et al, 2000). Bukchin extended his previous work in m/1/C to n jobs, but this time allowing interleaving (Bukchin, Masin & Kirshner, 2010). Many researchers studied the no-wait FSLS problems not allowing interleaving but integer sizes (m/N/C/DV) were assumed (Hall, Laporte, Selvarajah & Sriskandarajah, 2003;Kim & Jeong, 2009;Kumar et al, 2000).…”