Related Experiment Videos
The 'problem' with automation: inappropriate feedback and interaction, not 'over-automation'
1Department of Cognitive Science, University of California, San Diego 92093.
Summary
Automation design, not its presence, causes human error in high-risk industries. Improved automation design with better feedback and human interaction can significantly reduce errors.
Area of Science:
- Human-Computer Interaction
- Automation Engineering
- Industrial Safety
Background:
- Automation is increasingly adopted in high-risk industries.
- Current automation designs are often implicated in human error during failures.
- Existing systems provide inadequate feedback during abnormal situations.
Purpose of the Study:
- To identify the root cause of human error associated with automation.
- To propose design principles for mitigating automation-induced errors.
- To advocate for a shift towards more adaptive and interactive automation systems.
Main Methods:
- Conceptual analysis of automation design flaws.
- Examination of human-automation interaction during task control.
- Literature review on automation failures and human error.
Main Results:
- Automation's intermediate intelligence level is a key factor in failures.
- Inadequate feedback and interaction exacerbate human control difficulties.
- Current designs fail to account for all possible abnormalities.
Conclusions:
- Automation design must prioritize continuous feedback and human interaction.
- Systems should be designed to accommodate errors and worst-case scenarios.
- A move towards 'soft, compliant technology' is recommended over rigid systems.