Intelligent Alert Systems for Error and Conflict Detection in Supply Networks

Abstract Supply network collaboration requires seamless information integration to achieve good overall performance. Due to the huge amount of transaction data generated from distributed data sources, supply network information systems are highly error/conflict-prone. Detection of errors and conflicts in supply networks, however, is difficult for decision makers because of information overloading. Therefore, it is critical to develop a system that can automatically detect errors and conflicts among distributed data sources as early as possible. As a solution, this research proposes an intelligent agent-based alert system. First, typical errors and conflicts in supply networks are defined and classified. For automated error and conflict detection, the alert system consists of agents in three layers. To explain the effectiveness of the system, a business case of handling a capacity conflict among production, distribution, and sales is presented. During a five-month test with three products, the system detected 228 errors and conflicts, which might not be detected by human users. The system helps decision makers quickly indicate errors and conflicts and resolve them promptly to minimize the loss and enhance collaboration performance.