Display Design: Recent Technique Discussed

NEWSLETTER ARTICLE

Maintaining situational awareness, the "big picture" of unit operation, has been identified as a critical element in the ability of plant operators to handle process transients. The lack of situational awareness has been cited as a contributing factor in numerous operator error events across a wide range of industries, including Three Mile Island. Continual research is underway to identify more effective ways of transmitting data and improving situational awareness, while maintaining consistency with the human information processing system. One developing methodology, called Ecological Interface Design (EID) was recently applied to a simulated pasteurizing process and discussed in a recent Human Factors Journal article (Reising, D. and Sanderson, P., "Ecological Interface Design for Pasteurizer II: A Process Description of Semantic Mapping," Human Factors, Vol. 44, No.2, Summer 2002, pp.222-247).

EID is distinct from other interface design methodologies in that it relies heavily on abstraction principles as well as skill/rule/knowledge-based behavior. From a human perspective, skill- and rule-based behaviors require less mental process capability than knowledge-based behavior. In other words, less thought is required; processing occurs at a more subconscious level. With EID, critical factors in a process are identified and presented in a manner seeking to encourage use of skill- and rule-based behavior, while supporting knowledge-based behavior. This manner is often abstract in nature, with multiple variables integrated to provide higher level information to the operator.

An example of presentation is shown in the accompanying figure of an ecological display resulting from the EID process. While there are many aspects to the display that are beyond the scope of this discussion, one feature stands out. An emergent s-shape (shaded for this discussion) has been incorporated to alert the operator to changes in the process. When this shape varies, it is an indication that something in the process has changed, requiring operator attention. This rule, "s-shape = normal process," becomes one standard by which the process is monitored. From a human processing standpoint, identifying when this rule is broken is easier than monitoring different numeric parameters and knowing when they have varied from normal values (knowledge-based behavior). EID uses principles such as these to provide "big picture" awareness to the users.

While most operators preferred more traditional, "mimic" displays (based on process flow), Reising and Sanderson found that performance with ecological displays did improve in fault detection. However, performance varied widely among participants, with some never grasping the ecological displays.

In his review of EID, Vicente found that the inability to understand ecological displays is a common occurrence (Vicente, K.,"Ecological Interface Design: Progress and Challenges," Human Factors, Vol.44, No.1, Spring 2002, pp. 62-78.) He suggests that these variances result from differences in cognitive style of the individuals. People that are "serialists" tend to focus on details and individual facts, whereas "holists" are system thinkers and focus more on the big picture. Holists tend to perform better with ecological displays.

What conclusions can we draw about this new methodology? First, it needs to be noted that as of the writing of this newsletter, these displays have not been tested in industry. Trials thus far have occurred in simulated environments. Second, if these displays are successfully applied to industry, there may be implications for operator selection. Increased testing and training to produce qualified individuals may be necessary. Finally, while new methodologies are continually arising, caution should be exercised in their application. Was usability testing incorporated in the design process? What were the results? Where has this approach been implemented? Have they seen marked improvement in operator assessment of unit status? Questions such as these need to be asked to ensure that although the new graphics look impressive, they actually result in better human-system performance.

Copyright 2003 Beville Operator Performance Specialists, Inc., All Rights Reserved