This half-day workshop follows a taxonomy of real-world HRI, developed based on the PubRobFails and ExactingHRI workshops, as well as the HRI study life cycle tool. The taxonomy structures real-world HRI into phases such as design, evaluation, analysis, and reporting, and highlights common challenges and failure points across these stages.
The workshop opens with a short introduction to the taxonomy and an ice-breaking activity. The participants are then organized into group discussion sessions following the taxonomy, during which they present short stories of deployment challenges and failures from their experience. In this discussion, they both critique the taxonomy on its applicability and coverage, such as missing factors or methods, as well as discuss design and methodological revisions that would benefit their presented examples. This allows the groups to work on guidelines focusing on themes such as technical/environmental failures, human factors, ethics/governance, and data/evaluation.
This workshop is intended for HRI researchers, roboticists, designers, UX professionals, as well as those working in application domains such as healthcare, disaster response, industrial inspection, service robotics, and public-space deployments --- anyone who deploys robots in the wild and faces the complexities of real-world interaction. This workshop will be held in person only: given the sensitive and reflective nature of the discussions, particularly around failure, lessons learned, and behind-the-scenes realities, a shared physical space is essential to foster open, candid, and constructive conversations.
Our workshop will serve as a platform that allows researchers to openly discuss their experiences, including failures and the lessons they learned from them. Our goal is to create a collaborative space for sharing experiences, addressing challenges, and developing innovative approaches to tackle failure in research, promoting transparency, resilience, and innovation.
The workshop will generate both practical resources and community-level benefits:
Documentation and tools: Checklists, guidelines, and a practical toolkit for planning, executing, and analyzing real-world HRI deployments.
Cross-domain synthesis: A taxonomy of deployment failures and lessons learned that spans social, public, and extreme environments.
Community building: A strengthened community of practice across social and field robotics, normalizing the sharing of failures and lessons.
Future impact: A foundation for recurring workshops and collaborations on real-world HRI challenges.
Tentative agenda (in local time) for this half-day in-person workshop:
08:45-09:00 Taxonomy overview Welcome and overview to the real-world HRI taxonomy by organizers
09:00-09:15 Ice-breaking: Meme challenge A short, interactive activity where participants share or create memes (images, quick sketches, or captions) that capture the gap between expectations of robot performance and the messy reality of real-world deployments
09:15-09:45 Phase 1: Design Research questions and hypothesis; physical, social, locational, temporal, informational context; robot and sensor choice
09:45-10:15 Phase 2: Evaluation Quantitative, qualitative, and mixed methods; recruitment and data collection scale and approach; mitigation and toubleshooting; participant and stakeholder management
10:15-10:45 Break Morning tea
10:45-11:15 Phase 3: Analysis Data cleaning and processing; annotation; statistical analysis
11:15-11:45 Phase 4: Report Data / code open-sourcing and benchmarking; Follow-up studies; Publication
11:45-12:15 Summary Discussion summary and closing remarks by organizers