Wafer fabrication in display manufacturing is a highly complex and dynamic re-entrant process in which wafers go through hundreds of processing steps, entering the same processing group of machines several times. The success of display manufacturer is determined by its ability to provide the quantity and quality demanded by customers in an extremely competitive environment. For success of display manufacturing, they have to improve flow time, WIP level, throughput rate and equipment utilization. Assigning lots to the appropriate equipment can improve these performance measurement factors.

 There are several ways to assign lots to machine and one of them is the machine priority based lot assignment method. This method involves ranking machine in each machine group and assigning lots to the machine in order of priority. The priority of each machine is determined based on its current status such as idle, process etc.

 Unnecessary detour can cause by applying priority to each machine. If we assume that the priority of the PTH1 machine is higher than that of the PTH2 machine, PHT1 will pick up the closer lot, AA03, and AA01 will be assigned to the PHT2 machine. It can cause unnecessary detour compared to when AA01 is assigned to the PHT1 machine and AA03 is assigned to the PH2 machine. 

Fig 1. Example of unnecessary detour

1. Heuristic algorithm

The fisrst approach is to use heuristic algorithm. For example, assigning the lot closest in distance to a particular machine or the lot with the estimated earliest arrival time. Altough heuristic algorithms enable simple and fast computation, the solution obtained may not be the optimal solution.

2. Mathematical model

Using mathematical model is another solution to approach lot assignment problem. Objective function can be minimize distance, waiting time, transfer time or maximize machine utilization, total throughput. The approach using mathematical models can obtain the optimal solutionm, but it has the disadvantage of taking longer computation time and not considering dynamic situations.

[1] M.Vimala Rani, “A state-of-art review and a simple meta-analysis on deterministic scheduling of diffusion furnaces in semiconductor manufacturing”, 2021

[2] Juan Liu, et al., “Scenario-based multi-objective robust scheduling for a semiconductor production line”, 2019