Understanding operator decision-making is a good first step in improving operator effectiveness. Operator decision-making depends on the person (their level of expertise) and the situation (how familiar). A popular behavioral model from Rasmussen proposes that operator response can be broken into three levels; skill-based behavior, rule-based behavior, and knowledge-based behavior as shown in the figure below. 

Figure. Rasmussen’s SRK Levels of Cognitive Control

For all three levels, the need for a response is prompted by some form of sensory input (alarms, process data changing on the HMI, alarm horn, sound made by process equipment). The input undergoes different cognitive processing depending on the operator’s experience level and familiarity with the situation.

An operator who is extremely experienced with the situation tends to process the information at a skill-based level, reacting and responding automatically at a subconscious level. This is similar to a driver instinctively turning the steering wheel, without thinking about it, when their car starts to veer off the road.  Sometimes called Level 1 thinking, skill-based behaviors require minimum attentional resources.

An operator who is familiar with the situation, but does not have extensive experience with it, would process the input at a rule-based level. They compare the inputs to rulesets that they had accumulated from past experience or to written checklists / procedures. Rules can be thought of as “if-then” associations between inputs and appropriate actions. An example would be the operator detecting a low flow rate into a tank and increasing valve output so that flow rate reaches its setpoint. Providing feedback to the operator via the HMI and alarm system will help them improve their response. Those who don’t receive feedback will become more confident in their response, but not necessarily better.

When the situation is new and the operator has no previous experience with it, they process the input at a knowledge-based level.  They have no rulesets for comparison, so they would need to be much more deliberative in their thought process, which would require significant attentional resources. An inexperienced operator will go back to their conceptual understanding and mental model of the process to diagnosis the situation and develop a plan for taking action. An example would be the operator detecting a high feedstock flow rate into a tank that, combined with the other flows into the tank, would result in an overflow in 15 minutes if not reduced. 

Novices (new operators) work mostly at the knowledge-based level and sometimes at the rule-based level with the help of written procedures. As they become more experienced and as a result of training, they may operate more at the rule-based level. The expert (experienced operator) tends to work at the skill-based level most of the time, but will move between the three levels depending upon the situation. 

Effective operator decision-making depends on designing the alarm system and HMI to support all three levels. The likely types of human error are significantly different for each behavioral level as shown below. Errors in rule-based and knowledge-based behavior, which are called mistakes, are common to inexperienced operators. Skill-based response errors, such as slips and lapses, are typically made by experienced operators. The techniques for preventing human error are thus different depending upon the operator’s behavioral level.

Figure. Types of Human Error and Their Characteristics

For more discussion on the human factors behind operator response to abnormal situations, check out the webinar: Impact of Human Factors on Operator Response to Alarm: Lessons from Process Industry Incidents

Or the previous blogs about Human Error

• https://www.exida.com/Blog/know-your-human-error-part-i
• https://www.exida.com/Blog/know-your-human-error-part-ii

References: 

Kahneman, “Thinking Fast and Slow”, Farrar, Strauss and Giroux, 2011.

Klein, “Sources of Power: How People Make Decisions”, the MIT Press, 1999.

Rasmussen, “Skills, rules, and knowledge - signals, signs, and symbols, and other distinctions in human performance models”. IEEE Transactions on Systems, Man, and Cybernetics. 13(3), 257-266, 1983.

Ritter, Baxter, Churchill, “Foundations for Designing User-Centered Systems”, Springer-Verlag, 2014.

Wickens, Lee, Li, Becker, “An Introduction to Human Factors Engineering, 2nd Ed.”, Pearson Education Inc., 2004.