Designing with Fallibility: Examining the Knowledge Politics of Agency, Methods, and Motivations in Robot Failure Research

Paper Title

Designing with Fallibility: Examining the Knowledge Politics of Agency, Methods, and Motivations in Robot Failure Research

Publication Info

• Topic area: Examining the sociotechnical, ethical, and political dimensions of robotic fallibility in design research.
• Keywords: robotic fallibility, human-robot interaction, feminist technoscience, design research, epistemology, failure, sociotechnical systems, HRI, HCI, critical design.

Background and Problem

• Problem / challenge: Existing research on robotic fallibility often frames failures as opportunities for improvement but neglects the social, ethical, and political dimensions of failure. There is limited critical reflection on how failure is defined, studied, and addressed, as well as whose perspectives are prioritized.
• Significance: Understanding robotic fallibility as a sociotechnical phenomenon can lead to more inclusive, reflexive, and politically aware design practices, addressing broader societal implications of human-robot interaction.
• Motivation and related work: Prior work in HCI and HRI has explored failure as a design opportunity, focusing on user trust, interaction quality, and system robustness. However, these studies often privilege designers' perspectives and overlook collective, relational, and contextual dimensions of failure. This paper builds on feminist technoscience to critically examine these gaps.

Solution

• Proposed approach: A feminist technoscience framework to analyze robotic fallibility, focusing on three key dimensions: agency (who defines failure), the performativity of methods (how methods shape real-world outcomes), and research motivations (why failure is studied and for what purposes).
• Novelty:
  1. Reframes robotic fallibility as a socially embedded and contested category.
  2. Highlights the performative role of research methods in shaping robotic failure in real-world contexts.
  3. Critiques the motivations behind robotic fallibility research, emphasizing the need for reflexivity and inclusivity.
  4. Provides design recommendations for more ethical and socially responsive approaches to failure.
• Procedure and key techniques:
  • Conducts a critical interpretive review of robotic fallibility literature in HRI and HCI.
  • Applies feminist technoscience concepts to analyze how failure is framed, studied, and addressed.
  • Synthesizes findings into actionable design recommendations.

Results

• Concrete findings:
  • Designers are often privileged as the primary interpreters of robotic failure, while users are reduced to passive data sources.
  • Research methods not only study but also perform and shape robotic fallibility in real-world contexts.
  • Motivations for studying failure often prioritize user experience and economic gains, potentially neglecting broader ethical and societal concerns.
• Advantage over baselines: Provides a critical, theory-driven perspective that goes beyond technical and behavioral analyses, addressing the sociopolitical and ethical dimensions of robotic fallibility.
• Experiments / evaluation:
  • The paper conducts a critical review of HRI/HCI literature, tracing citation networks and focusing on studies where failure is central to the research question.
  • Examines methodological setups, epistemic assumptions, and real-world implications of robotic fallibility research.
• Limitations and future work:
  • The analysis is interpretive rather than systematic, focusing on specific epistemological and ethical questions.
  • Calls for further empirical studies on collective, relational, and emotional dimensions of robotic fallibility.
  • Suggests exploring the implications of "unmaking" and "undesigning" as alternatives to repair-focused design.

Summary

This paper critically examines how robotic fallibility is framed, studied, and addressed in HRI and HCI research, using a feminist technoscience lens. It highlights the social, ethical, and political dimensions of failure, emphasizing the need to move beyond designer-centric and economically driven approaches. The authors argue for recognizing failure as a relational and contested phenomenon, shaped by methodological choices and research motivations. They propose design recommendations that prioritize reflexivity, inclusivity, and care, encouraging researchers to rethink the normative assumptions underlying robotic fallibility studies. This work contributes to critical HRI by advocating for more socially responsive and ethically grounded design practices.