From Monitoring to Attention Economy: Event Design for Human Supervision in Highly Automated Industrial Systems

Industrial production plants in sectors such as chemicals, life sciences, metals and mining, pulp and paper, and power generation operate with a high degree of automation. During steady-state operation, control systems maintain process stability with little need for human intervention, while operators primarily supervise plant behavior and intervene during startup, shutdown, product transitions, or abnormal situations. At the same time, modern plants must monitor multiple domains simultaneously, including process performance, asset conditions, cybersecurity threats, and physical security. These developments significantly increase the number and diversity of events generated by monitoring systems. Advances in sensing technologies, connectivity, and data analytics have greatly improved the ability to detect deviations from normal operation. However, the increasing volume of detected anomalies and notifications raises the challenge of determining which events should be brought to the attention of human supervisors. Since direct human intervention becomes less frequent but more critical in highly automated plants, the effective use of operator attention becomes an essential aspect of system design. This paper examines monitoring within the broader framework of supervisory control in industrial systems. Monitoring is viewed as the continuous determination of system state across multiple domains, with anomaly detection forming the basis for subsequent supervisory decisions. Building on this perspective, the paper analyzes how detected deviations are translated into events and discusses criteria for deciding which events require human attention. A structured view of monitoring and supervision is proposed that integrates process operation, asset health, cybersecurity, and physical security. The goal is to outline design principles for event generation that support effective human oversight in increasingly autonomous industrial environments.