A hierarchical system for effective coordination of available-to-promise logic mechanisms

Abstract

This paper aims to provide a combinatory approach towards addressing the advanced available-to-promise (ATP) problem, consisting of three deterministic optimisation models that operate on both sides of the Customer Order Decoupling Point. The proposed approach is based on long-term aggregate capacity reservation for periods when increased volatility is expected, while still obtaining production plans that meet the predefined and agreed customer service levels. The three optimisation models together guide a system that helps manufacturers to optimally decide on ATP quantity and due date quoting on the basis of available manufacturing resources. To support this system, a prototype software module was designed and implemented in Java that loosely integrates with the popular Open Source ERP system Compiere2's databases and uses the Linear Programming solver QS-Opt to solve the models developed in this research. The system response times as evidenced in the experiments described in this paper are quite acceptable for real-world operations. The proposed solution of the ATP problem is of great value for all competitive and proactive organisations that need a practical tool to support, in the best possible way and in an almost real-time fashion, their decision on whether to accept or decline an incoming customer order request. It is our belief that an integration of the proposed models into existing ERP systems will enhance their limited ATP functionality and provide management with a powerful decision support tool.