An intelligent multi-constraint finite capacity-based lot release system for semiconductor backend assembly environment

Abstract

Listen

Translate article

This paper presents an intelligent multi-constraint finite capacity-based lot release system which has been designed, developed and implemented to solve the lot release problems in a discrete manufacturing environment with huge product mix and multitude of capacity constraints. This system releases lots by firstly prioritising them according to a multi-attribute lot prioritization criteria, and then by applying predefined multiple capacity constraints to the sorted queue of lots. Finally the finite resource capacity is allocated to them sequentially. This system is designed to be adaptable and re-configurable in response to the dynamic changes in the manufacturing environment. The system can be configured to operate in both an automatic and manual lot release modes. When the system is triggered under the automatic lot release mode, it will perform lot releases automatically based on the latest lot orders information and the predefined configuration, parameters and capacity constraints. It facilitates the auto generation of lot release schedule during night shifts when the production control personnel is off duty, as the system can be triggered through scheduled task provided by the computer's operating system. The manual release mode is designed to cater for fine adaptation of the lot release schedule generated by the system through potential human intervention, through which, a user is allowed to adjust the lot release schedule, in response to some last minute urgent requests from the customers and unexpected events from the shop floor. Detailed reporting tools that provide a snapshot of the lot release results controlled by the various capacity constraints are available for analysis and further refinement of the lot release schedule. This system has been successfully implemented in a few Semiconductor Backend Assembly companies, and in most of these implementations, with direct integration to their existing manufacturing systems such as the Enterprise Resource Planning (ERP) and Manufacturing Execution System (MES).