The Automator: intelligent control system monitoring

A large scale control system may contain several hundred thousand control points which must be monitored to ensure smooth operation. Knowledge of the current state of such a system is often implicit in the values of these points and operators must be cognizant of the state while making decisions. Repetitive operations requiring human intervention lead to fatigue, which can in turn lead to mistakes. The authors propose a tool called the Automator based on a middleware software server. This tool would provide a user configurable engine for monitoring control points. Based on the status of these control points, a specified action could be taken. The action could range from setting another control point, to triggering an alarm, to running an executable. Often the data presented by a system is meaningless without context information from other channels. Such a tool could be configured to present interpreted information based on values of other channels. Additionally, this tool could translate numerous values in a non friendly form (such as numbers, bits, or return codes) into meaningful strings of information. Multiple instances of this server could be run, allowing individuals or groups to configure their own Automators. The configuration of the tool will be file based. In the future these files could be generated by graphical design tools, allowing for rapid development of new configurations. In addition the server will be able to explicitly maintain information about the state of the control system. This state information can be used in decision making processes and shared with other applications. A conceptual frame work and software design for the tool are presented.