Frontier stage

Beyond one-size-fits-all: AI and social robotics for assessing child mental wellbeing

In person
  • Date
    10 July 2025
    Timeframe
    10:40 - 11:00 CEST
    Duration
    20 minutes
    Share this session

    How can we trustworthily and effectively support children’s mental wellbeing amid stretched clinical resources and societal stigma? This talk presents an interdisciplinary vision leveraging social robotics and AI to transform mental wellbeing assessment for future generations, developed through a collaboration between the University of Cambridge’s Affective Intelligence and Robotics Lab and Department of Psychiatry.

    Through studies with children aged 8-13, we designed robot-led assessments that combine structured interactions, validated questionnaires, and AI models. Our findings show that robotised assessments identify wellbeing concerns more effectively than traditional self- or parent-reports, with remote longitudinal interactions also proving effective. Automated analysis of children’s nonverbal behaviours revealed that children with higher wellbeing tend to express themselves more openly, with notable gender differences. These insights challenge one-size-fits-all approaches and highlight the need for personalised assessment tools that account for individual differences.

    We must adapt and revalidate psychological measures for digital delivery, as many tools were never designed for AI or robotic contexts. While Vision-Language Models (VLMs) show promise in interpreting children’s narratives, they also reveal bias underscoring the need for thoughtful, trustworthy integration of AI in child mental health context.