This page provides a very brief overview of the outcomes of the workshop. During the workshop we focussed on the challenges in these fields, but we also tried to think of solutions to these challenges.
Summary of CRITTER 2023
The field of child-robot-interaction (CRI) is a fast growing subfield of human-robot-interaction (HRI), with a multitude of potential contexts and applications. Designing and conducting successful CRI studies often requires input from several different areas, necessitating interdisciplinary collaborations. However, establishing such collaborations is not always straightforward, and can lead to difficulties with consolidating definitions, measurement techniques, and experimental design. For this reason, a workshop with researchers from different subfields and application domains of CRI was organized at the HRI 2023 conference in Stockholm to identify the key challenges and potential solutions of interdisciplinary research. The workshop opened with a discussion about interdisciplinary research among panelists with expertise in developmental psychology, interaction design, robotics and control, and computer science. After that, 7 CRI-related flash talks were presented, introducing starting points for the 3 following breakout sessions. During these sessions, participants of the workshop split into groups and identified key challenges in the field of CRI and potential solutions to these challenges. In the end, these insights were shared and discussed with the entire group. Below is a summary of the most important insights of the workshop. We hope that some of these ideas will encourage both current and future CRI researchers to think critically about the design, measurement, and long-term implications of child-robot-interaction, and to work together towards developing and sharing resources which can benefit the community as a whole.
Opportunities for interdisciplinary research
- Reproducibility. To create valid (open source) measures, guidelines, checklists, protocols and good software practices. In light of this point, a specific suggestion was made to organize smaller (regional) networks that share their insights and practices on a larger scale (while also leaving space for unique ideas from outside collaborations).
- Robot differentiation. Especially when it concerns language. In an international team, they may anticipate that the robot can make distinctions, understand different languages/dialects.
- Content creation.Children easily get bored with the risk that they do not want to interact with the robot anymore. To engage children for a longer period of time, more content and more personalized content is needed. It is therefore important to share each other’s content and to collaborate with professional content creators.
- Attract funding. Interdisciplinary collaborations attract more funding, while also creating more visibility for researchers from low/middle income countries.
Challenges of interdisciplinary research
- Different perspectives and expectations due to different experiences. It is important to start the collaboration with a common ground. Related to this point, it was specifically mentioned that researchers often make the same mistakes. To evolve in CRI research, previous mistakes should be shared in order to learn from each other. At the same time, technology also develops over time. This difference in knowledge and experience can pose new challenges.
- Different concepts in the field. Researchers from different domains tend to have different understandings of academic terms, such as “engagement” or “relationship”.
- Finding agreement in goals & research questions. Having different interests and research agendas often makes it difficult to determine the common goals, focus and research questions.
- Finding agreement on authorship. A more practical point is that researchers find it challenging to determine authorship order based on contribution.
- Strict collaboration beliefs. Researchers can have strict beliefs that hinder collaborations, for instance, that a specific type of research only belongs within a specific research domain (e.g., psychology). They need to think more interdisciplinary and be open to learn from others.
Solutions of interdisciplinary research
- Determining the goals and research questions beforehand. To create common ground.
- Frequent communication. Having a lot of meetings beforehand (e.g., about the study design, metrics, goals, research questions, authorship, responsibilities) and during the collaboration prevents miscommunication and ensures that everyone’s interests and expertise are taken into account. An extra tip is to actually schedule various types of meetings together (e.g., hackathons, writing sessions, brainstorms).
- Writing down agreements. Be sure to record important points in writing. It is both a check to see if everyone has interpreted the decisions in the same way and a point of reference for later decisions and discussions.
- Trust. Make sure that there is trust in each other’s expertise, execution and accountability. If you are working with the best people, you more easily accept what they are doing (and the other way around).
Challenges and solutions of child-robot-interaction
- Long-term child engagement. Solutions: participatory design (incl. children’s preferences and perceptions), game-based activities, increased task difficulty, work with multidisciplinary teams (e.g., creatives, psychologists, educators).
- Managing robot expectations and responses to errors. Solutions: have the robot be clear about its limitations, be careful about designs that imply more functionality than it has, develop coping strategies for errors or long-term issues.
- Robot safety. Solution: create soft and less-destructible robots.
- Data privacy. Solution: inform children about safety (how results are being processed).
- Conflicting needs and goals stakeholders (e.g., “I want my kid to learn” vs. “I want it to do my homework”, “I want the robot in my room” vs. “It should be in a common space”). Solution: constantly communicate and collaborate with these stakeholders.