Population ageing and the growing number of older adults living alone are creating new challenges for health and social care systems. Early detection of changes in cognitive or functional status is essential to enable timely intervention before risk situations arise.
TEMPRIA is a research and development project that applies artificial intelligence to the non-invasive monitoring of daily activities at home, identifying changes in routine patterns that may indicate physical, cognitive or emotional decline.
Its goal is to support a preventive, person-centred approach to care while respecting the autonomy and privacy of older adults.
How does TEMPRIA work
The project integrates environmental sensors installed in the home to monitor different indicators of daily activity, including:
- Mobility within the home.
- Activity patterns.
- Resting and sleep habits.
- Acoustic characteristics of the environment.
The system does not use cameras or record identifiable conversations, ensuring the privacy of participants.
The collected data are pseudonymised and analysed using artificial intelligence algorithms capable of identifying behavioural patterns and detecting significant deviations from normal daily routines.
When the system detects potential anomalies, it generates information that can support professionals in assessing the need for preventive intervention.
Project Objectives
TEMPRIA's main objective is to develop an intelligent platform capable of detecting early signs of cognitive decline by analysing the everyday behavioural patterns of older adults.
Its specific objectives are:
- Develop artificial intelligence models to detect behavioural anomalies.
- Characterise activity patterns within the home.
- Integrate functional, clinical and environmental data to train predictive models.
- Build a knowledge base that links detected changes to potential neurodegenerative processes.
- Contribute to a more preventive, personalised and efficient home care model.
Pilot Project
The project has successfully completed its pilot phase with the participation of four older adults living independently in their own homes.
During this period, multiple sources of information were combined:
- Continuous environmental monitoring.
- Periodic neuropsychological assessments.
- Functional capacity monitoring.
- Recording of clinically relevant events.
This phase enabled the validation of both the technology and the analytical methodology, while also providing real-world data to train the artificial intelligence models.
The results show that the system successfully characterised each participant's functional behaviour patterns and generated a dataset that will support the further development and validation of the predictive models.
Results Achieved
The TEMPRIA pilot phase has validated the monitoring and data collection methodology in a real-world setting, confirming the feasibility of the proposed solution.
Key outcomes:
🧠 Methodology validation
The system was successfully tested in real home environments, integrating technology, clinical monitoring and professional expertise.
📈 Creation of a clinical-behavioural dataset
Environmental, functional, cognitive and clinical data were collected to build a pseudonymised dataset for training the artificial intelligence models.
🏠 Behaviour pattern characterisation
The project identified the participants' typical daily activity patterns, establishing a baseline for detecting future behavioural changes.
🤖 AI model training
The collected data provide the foundation for the continued development of predictive models capable of identifying potential risk situations at an early stage.
🤝 Validation of the collaborative model
TEMPRIA has demonstrated the value of combining biomedical research, artificial intelligence and social innovation through the collaboration between Suara Cooperativa, La Salle-URL and IDIBGI.
🚀 Preparing for future phases
The outcomes lay the groundwork for scaling the project to new care settings and further validating the technology with a larger group of participants.
Key Results
🏠 Participating households
👵 Participants
📅 Pilot duration
🧠 Neuropsychological assessments
📊 Dataset generated
🤖 Technology
4
4
12 + 12 month
Quarterly
Pseudonymised clinical, cognitive and behavioural dataset
Artificial intelligence and non-invasive environmental sensors
A project developed in collaboration
TEMPRIA is a collaborative project developed by:
- Suara Cooperativa, leading social innovation and the validation of the solution in real care settings.
- La Salle – Universitat Ramon Llull, responsible for the technological development and the artificial intelligence models.
- IDIBGI – Dr. Josep Trueta Biomedical Research Institute of Girona, providing the clinical and neuropsychological expertise required to scientifically validate the project's outcomes.
This partnership brings together research, technology and care expertise to develop an innovative solution with real social impact.
Funding
TEMPRIA is a Research and Development (R&D) project funded through public investment to drive innovative solutions that improve care for older adults and promote a more preventive, personalised and sustainable model of care.
