Distributed Diagnostics and Dynamic Reconfiguration Using Autonomous Agents

Technological, economic, and cultural changes that are driving the deployment of distributed intelligent devices across a broad range of industrial, commercial, commercial, and military systems. Sensors and actuators that previously provided a single function are being replaced with smart devices that are programmable, perform multiple functions and provide realtime operating information. For example a motor speed controller that historically maintained a setpoint motor speed may now include embedded logic for local closed-loop control, dynamically alter speed in response to changing pump conditions, provide real-time diagnostics, and optimize performance [Discenzo, Rusnak et al. 2002]. The automation of increasingly complex, critical, coupled systems can place a severe demand on centralized automation systems. A framework for managing this complexity employing intelligent agents has been demonstrated for automation systems such as package handling, material processing, and water distribution. The agent infrastructure may be extended to provide for distributed collaborative diagnostics of novel faults. Following fault detection, agents then collaborate to identify and implement a loss mitigation strategy, recovery strategy,