Conversational AI for digital mindfulness with older adults

Mindfulness has become a common self-care practice to improve mental well-being, decrease symptoms of stress and improve relaxation and concentration.  Mindfulness can be defined as a state that can be achieved through focusing one’s awareness on the present moment and environment (American Psychological Association). Mindfulness-based interventions are used as a prevention and treatment for depression, anxiety, stress and other psychiatric disorders, and they can be practiced through various exercises and assignments aimed at developing mindfulness through formal practice in daily life. 

This project aims to co-design a conversational AI mindfulness coach that provides personalised mindfulness exercises to improve the bodily and emotional well-being in older adults' daily lives. This will be achieved through a human-centric understanding of the use of a voice-guided digital mindfulness among older adults and integration of Large Language Models to a virtual mindfulness coach. Mobile applications have become common technology to collect data of user’s physical activity, nutrition and sleep, but the long-term user engagement with these technologies still remains low.  Conversational applications in the form of chatbots or virtual assistants can decrease many technology adoption barriers that older adults typically experience with digital technologies and improve the interactivity and user engagement with the device. Therefore, conversational interaction modalities provide a new possibility to develop interactive technologies for digital mindfulness that better meet the needs and interests of older adults.  

First, the project conducts a user study for a voice-based mobile application to understand older adults’ mindfulness practice, and identify development areas for improving interactivity of the application with older adults. The user study investigates older adults’ previous use of mobile applications in health monitoring, their perceived mental well-being, mindfulness and stress symptoms. 

Second, the project conducts a participatory design of a conversational application for digital mindfulness through developing a dialogue model for mindfulness-based interventions with older adults.  The project explores and develops the suitability of different interaction modalities (voice, sound, music) and Large Language Models as a part of digital mindfulness, and the user interface of the application with older adults.    

Research team:
Lucy McCarren, PhD student, KTH
Ulrika Eriksson, Research engineer, KTH 
Laura Ortiz Mengual, Master student, KTH 
Sanna Kuoppamäki, Assistant Professor, KTH 

Previous students:
Yuqi Sun, Master student graduate, KTH
Magnus Thulin, Master student graduate, KTH

Publications:

• McCarren, L., Kuoppamäki, S. (2025). Design Preferences, Routines, and Well-Being of Older Adults Using Voice-Guided Digital Mindfulness: Qualitative Interview Study. JMIR Human Factors (2025);12:e67533. DOI: 10.2196/67533
   
• McCarren, L., Kuoppamäki, S. (2025). A LLM-Based Chatbot for Mindfulness Practice with Older Adults: A Development and Usability Study. Studies in health technology and informatics. DOI: 10.3233/SHTI250493

Conference presentations and societal impact:

• Sanna Kuoppamäki, Lucy McCarren (2025). Digital mindfulness facilitating healthy ageing through the application of Conversational AI. MSCA EU Lullabyte Network meeting. 
• Lucy McCarren, Magnus Thulin, Sanna Kuoppamäki (2024). Exploring and Defining Bias in Conversational Medical AI Assistants with Older Adults: Socio-Technical Approach. Workshop presentation in Exploring Bias in Generative Models (WASP-HS Workshop in AI for Humanity and Society 2024)
• Sanna Kuoppamäki (2024). Teknik och AI i vård och omsorg: Fördelar och utmaningar för äldre vuxna. November 6, 2024. Östermalms Sjukvårdsförening, Stockholm. 
• Lucy McCarren, Sanna Kuoppamäki (2024). Enhancing older adults' well-being through converstional application for digital mindfulness. Nordic Sociological Conference, University of Linköping. August 14-16, 2024.
• Lucy McCarren (2024). The use of digital mindfulness to facilitate self-perceived bodily and emotional well-being among older adults. Socio-gerontechnology Network Annual Meeting. Technical University of Vienna, 19-20 September, 2024. 
• Sanna Kuoppamäki (2024). Relationality in dialogues with conversational AI companions and older adults. Socio-gerontechnology Network Annual Meeting. Technical University of Vienna, 19-20 September, 2024. 
• Lucy McCarren, Sanna Kuoppamäki (2024). Utvecklingen av en konversationsapplikation för mindfulnessträning med äldre vuxna. Välfärdsteknikmässan, Stockholm Stad. 26 maj, 2024

Master thesis:

• Laura Ortiz Mengual (2025). Age-Inclusive User Interface Design for an AI Mindfulness App (LugnAI). KTH, School of Engineering Sciences in Chemistry, Biotechnology and Health (CBH), Biomedical Engineering and Health Systems.
• Age- Magnus Thulin (2024). Investigating an Age-Inclusive Medical AI Assistant with Large Language Models: User Evaluation with Older Adults. KTH, School of Engineering Sciences in Chemistry, Biotechnology and Health (CBH), Biomedical Engineering and Health Systems.
• Yuqi Sun (2024). MeditAI: Fine-Tuning Pre-trained Large Language Models for Guided Mindfulness Practice. KTH, School of Electrical Engineering and Computer Science (EECS) 

Research communication:

• Exploring AI-mediated mindfulness practice to promote older adults' well-being (Blog post, 2024-02-13)
• Delta i mindfulness-forskning på KTH (Blog post, 2024-02-16)

Funding: KTH Innovation, Adaptive Intelligent Homes (Digital Futures)