Integrating Flow Theory and Adaptive Robot Roles: A Conceptual Model of Dynamic Robot Role Adaptation for the Enhanced Flow Experience in Long-term Multi-person Human-Robot Interactions

Authors: Huili Chen, Sharifa Alghowinem, Cynthia Breazeal, Hae Won Park · HRI 2024

Licences

Open Access
Paper: CC BY 4.0

Everything below draws on the full paper under its CC BY 4.0 licence; its figure is credited to the authors.

TL;DR

Set a group robot's adaptation goal as the users' flow experience, a balance of challenge and skill, and reach it by shifting the robot's support role between leader and follower.

Why it matters

Social robots in homes, classrooms and care settings increasingly interact with small groups over many sessions, where a fixed behaviour policy is hard to design. The model gives designers a structured way to decide what a group robot should optimise and which behaviour to change, while keeping the robot's social identity and personality fixed so they do not confound the interaction.

Problem

Robot adaptation work has mostly targeted single users with one objective that stays fixed for the whole interaction, and often adapts the content rather than the robot's own behaviour. Most multi-person HRI studies use Wizard-of-Oz teleoperation or rule-based policies that cannot adapt in real time. The authors find no generalized model for long-term adaptation that covers both single-person and multi-person interaction.

Proposed solution

The robot's adaptation goal is the users' flow experience, split into cognitive and affective sub-objectives at individual and group level, whose weighting is set by group factors (for example sociodemographics, attitudes to robots, baseline behaviour) and activity factors (progress, resources, environment) and can shift within or across sessions. The robot pursues this goal by moving along a support-role spectrum from follower, learner or tutee to coach, leader or tutor: more leader-like when group performance is the priority, more follower-like when interpersonal dynamics are.

Key insights

Flow has a cognitive dimension (activity challenge against user skill) and an affective dimension (for example rapport), and the two regulate each other, so a robot's objective needs both.
Group-level affect cannot be obtained by averaging members' affective states; constructs such as dyadic rapport, interpersonal synchrony and group cohesion are needed.
A policy that maximises group performance at the cost of low performers or unequal allocation of interaction resources is unlikely to give the group an optimal experience.
Role-switching frequency must suit the context: a sudden switch from demonstrator to help-seeker within one activity can confuse the group about what the robot knows.
Only the robot's support role should adapt; changing its social role, identity or personality would add confounding variables.

Main results

Defines a hierarchical adaptation objective with four sub-objectives: individual cognitive, group cognitive, individual affective and group affective, whose weighting is context-dependent and may change within or across sessions.
Lists four group-related factors (d.1 sociodemographics and relationships, d.2 perceptions of the activity and robots, d.3 baseline behaviour, d.4 state just before a session) and three activity-related factors (a.1 progress, a.2 resource constraints, a.3 context and environment) that set that weighting.
Replaces the three-category follower, peer and leader classification of robot roles in groups from Sebo et al.'s survey with a continuous leader-to-follower support-role spectrum, defined relative to the roles of the human members.
Formalizes dynamic robot role adaptation with two worked examples: a follower-like role when interpersonal dynamics dominate the objective, a more leading role when team performance does.
Reports no quantitative results of its own; the authors state the model has been applied in a separately published empirical study of parent-child story-reading with a robot.

Limitations

Offers no ready-to-use metrics or measurements of flow experience in multi-person HRI; turning the framework into measurable metrics is left to future work.
It is a conceptual model, not a practical design tool or system development framework for HRI practitioners.
Implies complex sensing and perception of group dynamics, emotions and performance that the authors say may be technically difficult with current autonomous HRI systems.
Needs further empirical validation through user studies, case studies and quantitative analyses.
The leader-follower support role is only one possible axis of behaviour design; other spectra, such as affective-cognitive or motivational-informational, may suit some contexts better.

Field impact

Winner of the HRI 2024 Best Paper Award (Theory & Methods track, full papers), with 21 citations. It is a conceptual paper and releases no code, data or hardware designs.

Abstract

In this paper, we introduce a novel conceptual model for a robot’s behavioral adaptation in its long-term interaction with humans, integrating dynamic robot role adaptation with principles of flow experience from psychology. This conceptualization introduces a hierarchical interaction objective grounded in the flow experience, serving as the overarching adaptation goal for the robot. This objective intertwines both cognitive and affective sub-objectives and incorporates individual and group-level human factors. The dynamic role adaptation approach is a cornerstone of our model, highlighting the robot’s ability to fluidly adapt its support roles—from leader to follower—with the aim of maintaining equilibrium between activity challenge and user skill, thereby fostering the user’s optimal flow experiences. Moreover, this work delves into a comprehensive exploration of the limitations and potential applications of our proposed conceptualization. Our model places a particular emphasis on the multi-person HRI paradigm, a dimension of HRI that is both under-explored and challenging. In doing so, we aspire to extend the applicability and relevance of our conceptualization within the HRI field, contributing to the future development of adaptive social robots capable of sustaining long-term interactions with humans.

Open Source Robotics