What Is My Robot Thinking? Design Considerations for Transparent and Trustworthy Shared Autonomy
This research paper investigates how interface-level transparency—specifically varying feedback modality (visual vs. auditory) and information richness (sparse vs. rich)—shapes user interaction, coordination, and trust in a vision-based shared autonomy system. Through a user study with 25 participants across two assistive manipulation tasks of differing complexity, the authors evaluate how providing feedback influences intent alignment and corrective interventions. The findings indicate that providing feedback significantly improves goal legibility and reduces corrective inputs, while trust depends on task-appropriate information exposure rather than maximal belief disclosure.
Summary and content fields are drawn from the document text and are pending editorial review.
Who issued it
Who it governs
Researchers and developers in human-robot interaction and assistive robotics seeking guidance on designing transparent interfaces for shared control systems.
What the document contains
The document covers the design and empirical evaluation of transparent interfaces for vision-based shared autonomy systems, comparing visual and auditory modalities across sparse and rich information levels. It addresses how these interface variations influence intent alignment, corrective interventions, user understanding, and trust across two assistive manipulation tasks.
The text does not provide regulatory compliance frameworks, legislative mandates, or long-term longitudinal evaluations of trust.
- Providing feedback significantly improves user-robot intent alignment and reduces the need for corrective interventions in shared autonomy systems.
- Exposing the robot's inferred goal is sufficient to enhance coordination, and users do not require full access to the internal belief distribution.
- Participants rated visual feedback as significantly more helpful for understanding and trusting the robot compared to auditory feedback.
- Preferences for feedback richness are task-dependent, where rich information helps in simpler tasks but can overwhelm users in cluttered environments.
- Auditory richness can increase cognitive load and distraction, suggesting that auditory cues should be limited to brief, sparse confirmations of intent changes.
· I. INTRODUCTION · II. RELATED WORK · III. PRELIMINARIES · IV. INTERFACE DESIGN · V. EXPERIMENTS · VI. RESULTS · VII. DESIGN CONSIDERATIONS FOR TRANSPARENT SHARED AUTONOMY · VIII. CONCLUSION AND FUTURE WORK · IX. ACKNOWLEDGMENTS
How it scores
Each score is backed by a quote verified against the document text. Editorial review has not yet confirmed them.
Accountability
0/5Clear responsibility for AI outcomes
None
screening score · 2026-10-23
Fairness
0/5Non-discrimination and equitable treatment
None
screening score · 2026-10-23
Human Dignity
0/5Respect for the inherent worth of persons
None
screening score · 2026-10-23
Human Oversight
2/5Humans remain in control of consequential decisions
Assistive robots operating under shared autonomy must balance user control with autonomous assistance.
Abstract
screening score · 2026-10-23
Privacy
0/5Protection of personal data and private life
None
screening score · 2026-10-23
Transparency
5/5Openness about how AI systems work and are used
We investigate how interface-level transparency–feedback modality (visual vs. auditory) and information richness (sparse vs. rich)–shapes interaction in a vision-based shared autonomy system.
Abstract
screening score · 2026-10-23
Enforcement
0/5Consequences and compliance mechanisms with real force
None
screening score · 2026-10-23
Impact Assessment
0/5Required assessment of an AI system's risks/impacts before or during use
None
screening score · 2026-10-23