Influence of human-machine interactions and task demand on automation selection and use

Area of Science:

• Human-Computer Interaction
• Automation Systems Engineering
• Cognitive Psychology

Background:

• Established 10 levels of automation (Sheridan & Verplank) and a four-stage human information processing model.
• Introduced human-machine cooperation-centered models and cooperation modes.

Purpose of the Study:

• To empirically evaluate which human-machine theories best describe automation selection behavior.
• To determine the primary factors influencing user choices among different automation types.

Main Methods:

• Participants repeatedly selected from four automation types (no automation, warning, co-action, function delegation).
• Tasks involved completing three multi-attribute task battery scenarios, simulating recreational aircraft pilot operations.
• User choices were analyzed against human-machine cooperation and technology-centered models.

Main Results:

• Automation selection was driven by the perceived quality of human-machine interaction, not by automation performance.
• Participants favored automation types that offered superior cooperation, irrespective of their technical sophistication.
• Technology-centered models failed to accurately predict automation selection, unlike cooperation-based theories.

Conclusions:

• Human-machine cooperation quality is a more significant predictor of automation use than mere effectiveness.
• Future automation design should prioritize intuitive and high-quality human-machine collaboration.
• User-centered design principles are crucial for effective automation implementation in complex systems.