Blog

In systems with higher levels of automation (levels 6, 7, & 8), who has the most control over system operations and task execution? 

Consider a nuclear power plant operator working a morning shift (3am – 12pm) and an automated system that manages the plant. The automation integrates information from multiple places within the plant and provides alerts to the operator if there are values out of range. When the afternoon shift operator arrives at 12pm, they notice the temperature in reactor 1 is high and asks the morning operator how long the temperature has been rising. The morning operator replies that they have not received any alerts about the temperature in reactor 1 and was not aware that the temperature had been rising. The morning operator mentions that when they checked the temperature at 4am it was normal, and they were not aware it had changed. System diagnostics reveal that the temperature had been steadily rising since 5am but was not yet high enough to trigger an automated alert. Within hours, reactor 1 experiences a partial meltdown. What source of automation intervention failure is this describing?   

Which of the following is true of automaticity and highly automated tasks (that humans perform)?

When designing for long term memory, how can we support correct mental models? 

Consider the knowledge, rules, skills scale, and expertise in decision making. If a decision maker has no automatic response or rules to rely on and goes through an analytical process to make decisions, what level are they operating on? 

Fill in the blank…if you must perform an action to a high level of accuracy, you must act BLANK. 

In voluntary task switching, we discussed five factors that determine the attractiveness of a task. In this context, what does time on task refer to? 

We know that associations in long-term memory can impact decay and ease of retrieval. What else impacts memory decay or ease of retrieval? 

Consider an automated system being used in an operating room; this system has a “Quick Quiet” function so alerts can be instantly silenced if an emergency occurs with the patient until physicians are ready to reactivate the system. We discussed some examples of things humans are better at vs. things machines are better at. Why is it important that the system can be silenced/put in the background during an emergency? 

When we design for long-term memory, what is an important consideration?