In the Netherlands alone, dementia affects approximately 290K people. As the population ages and caregiver shortages grow, supporting people with early-stage dementia to maintain their ability to live at home as long as possible becomes crucial. In this project we aim to develop an intelligent and conversational home system to enhance the quality of life for elderly people with early-stage dementia. We propose to use Artificial Intelligence (AI) techniques to provide personalized care by mapping the need and preference of individuals.

Built based on the just-in-time adaptive intervention (JITAI) design concept, the conversational system will adapt to contexts of individual users and therefore provide the right types of support at right timings taking account personal preferences and contextual status.

We will design and develop multimodal interventions (combining text, image, music, etc.) for target users, as well as the optimal deliver strategies on conservational interactions between the intelligent system and human users.

The project will fit under the umbrella of KIA social theme ‘Health & Care’. This innovation enables families and caregivers to leverage advanced AI technology for improved support, offering societal and economic benefits by enhancing care quality and easing the burden on caregivers.

CALL
Wanted: participants for this research, november 2025!

Project duration

1 Jul 2025 -30 Jun 2028

Project manager

Shihan Wang, Utrecht University

Researchers

prof. dr Somaya Ben Allouch
Stephanie Kramer MSc
dr Michel Oey
Wen-Tseng Chang, Utrecht University

Funding

CLICKNL

Partners

illi Engineering B.V.

Until now the existing health and exercise apps in app stores lack a scientific base. Different apps use different ways to stimulate people to be physically active (e.g. feedback about physical activity, motivational messages, games). Some researchers argue that what type of app works, is highly individual. This means that the right type of app may depend on e.g. your current physical activity level, health, personality and residential context. However, how to make the optimal match between app and user is still unclear.

This project aims to answer these questions, by developing an exercise app with different forms of feedback. We will also develop a method to find the optimal match between app and user. We will use the app to collect data about the physical activity and location of the users. Since these data is collected every 5 seconds, this will give us very large data sets (big data). We will use ‘data mining’ techniques (automated search programmes on a computer) to find out which app works best for different types of users. We will then give each person his or her optimal app, and measure if the physical activity further increases. Based on the data, we can determine what the effect is of different types of exercise apps, and what app is most suitable for a particular person.

Image & Video

Project duration

1 May 2017 -1 May 2020

Project manager

Marije Baart de la Faille-Deutekom

Researchers

prof. dr Ben Kröse

dr Shihan Wang

Nicky Nibbeling

Monique Simons (Wageningen University)

Prof. Dr. Victor Zuniga Dourado (UNIFESP, Sao Paolo)

Karlijn Sporrel (Utrecht University)

Remi de Boer

Funding

Nationaal Regieorgaan Praktijkgericht Onderzoek SIA RAAK-mkb

NWO (Nederlandse Organisatie voor Wetenschappelijk Onderzoek)

FAPESP (Sao Paolo research Foundation)

Partners

• UvA – University of Amsterdam
• UNIFESP- Federal University of Sao Paulo
• UU – Utrecht University
• UFSCar – Federal University of São Carlos
• INFO
• IoT Living Lab BV
• ZappYou
• Municipality of Amsterdam
• Municipality of Utrecht
• HP2consulting