Increasing the Effective Span of Control: Advanced Graphics for Proactive, Trend-Based Monitoring

Abstract

OCCUPATIONAL APPLICATIONS Legacy status display formats limit the number of processes that human operators can effectively and proactively monitor. Through a design approach targeting the needs of human attention, a novel trend–icon hybrid status display format (Trendicon; patent pending; Fisher Rosemount Systems, Inc., Round Rock, TX) was developed for proactive monitoring. An experiment contrasting Trendicons with digital values and trend graphs revealed that after minimal training, Trendicons supported earlier problem detections and more frequent proactive detections. Further, Trendicons led to a roughly fivefold increase in process indicators that participants could effectively oversee. Practical implications include earlier problem detection, improved system operation, and an increased operator span of control. Participants’ incorrect initial intuitions about display effectiveness highlight potential pitfalls of flexible, configurable displays. Trendicons are being developed for process and military control displays. The attention management design approach taken here is applicable to the design of supervisory status views for display-mediated work domains. TECHNICAL ABSTRACT Background: Process control operators rely heavily on their visual displays to monitor and stay abreast of processes and situations, anticipate future status, and proactively detect and resolve emerging problems. However, fielded status displays support only a subset of these tasks and can engender reactive rather than proactive monitoring because of inadequate attention management. Reactive monitoring is a significant barrier to effective operations and increasing operator span of control. Purpose: The benefits of an attention management design approach are demonstrated through the development and evaluation of a novel trend–icon hybrid (Trendicon) display format to support proactive monitoring. Methods: A within-subjects experiment contrasted the effectiveness of three status indicator formats—digital values, trend graphs, and Trendicons—in supporting proactive detection of emerging deviations. Span of control was assessed by varying the number of concurrently monitored status indicators (4, 12, and 24). Participants’ intuitions and preferences for the indicator formats were assessed as a proxy for display configuration choices. Results: Monitoring was largely reactive for digital values and trend graphs. In contrast, Trendicons promoted earlier problem detections and more frequent proactive detections, leading to a nearly fivefold increase in effective span of control. However, participants initially underestimated the utility of Trendicons. Conclusions: By explicitly designing attention management into new representations of process status, a dramatic increase in effective span of control was achieved over legacy formats. Better allocation of limited attentional resources during monitoring helped shift the nature of monitoring from reactive to proactive. Miscalibrated performance intuitions and preferences demonstrate that designers must consider both performance and user beliefs about display effectiveness. Development of Trendicons is ongoing in process and military control display applications. Designing for and addressing the underlying cognitive challenges of display-mediated supervisory control can result in more effective system operation, safer management of abnormal situations, and increased productivity.