Reliance on Haptic Assistance Reflected in Haptic Cue Weighting

Area of Science:

• Human-Computer Interaction
• Automation and Control Systems
• Cognitive Psychology

Background:

• User trust in automation is crucial for performance in shared control systems.
• Reliance, the behavioral aspect of trust, significantly influences how users interact with automated systems.
• Prior research focused on supervisory control, but less on shared control dynamics.

Purpose of the Study:

• To investigate how reliance on haptic assistance affects performance in shared control.
• To quantify user reliance by manipulating the variability of haptic cue errors.
• To understand the impact of explicit versus implicit knowledge of automation errors on reliance.

Main Methods:

• Participants performed reaching tasks towards a hidden target using visual and haptic cues.
• Reliance was measured by analyzing endpoint positions relative to visual and haptic cues.
• Trial-by-trial random error variability in haptic assistance was systematically altered.

Main Results:

• Increased haptic error variability led participants to shift reliance towards the visual cue, demonstrating cue weighting.
• This shift in reliance occurred regardless of whether participants had explicit or implicit knowledge of the errors.
• Participants with explicit knowledge adapted their reliance more rapidly than those with implicit knowledge.

Conclusions:

• User reliance on haptic assistance in shared control is sensitive to the variability of automation errors.
• Explicit knowledge facilitates faster adaptation of reliance compared to implicit knowledge.
• The study introduces a quantitative method for assessing user reliance in shared control automation.