2022 - 2024

Ref. TED2021-130296A-I00

TAICare: Intelligent transformation of the ADL-based caregiving process and its intervention with digital solutions

In this project, Conversational Assistants (CAs) and/or Connected Devices (CDs) at home allow us to acquire relevant data from users’ interactions to subsequently analyze, with artificial intelligence (AI) approaches, abnormal behaviors at early stages when performing ADL tasks (which may lead to possible cognitive impairment). Besides, related digital solutions can act on daily living to support the long-term caregiving delivery to improve people life quality in a more sustainable and innovative fashion.

Overview

Main Objective: 

Support the long-time caregiving delivery for elderly people at home in order to improve their life quality and that of their carers, making use of sustainable innovative technologies and smart procedures.

To satisfy the proposed objective and go more in depth in its study, we also describe several contirbutions: (1) ADL-based multi-modal data acquisition framework, (2) Validated tools for ADL intervention and analysis of primary users (based on knowledge extraction), (3) Open datasets on low/high-level monitoring from acquired data (primary users) (4) Real change and impact on the digital transformation of caregiving process and ADL measurement scales.

Work Packages

Coordination, Management, Dissemination

Transversal aspects of the project. Ethical and privacy implications.

ADL acquisition framework for data monitoring

Model for data acquisition and monitoring ADL performance for behavior detection

Knowledge extraction and analysis

Data processing and knowledge extraction for ADL correlation and patterns detection

Pilot Trials

Field trials in real environments with seniors living alone

Research Team

  • José Bravo (UCLM, Computing Engineer)
  • Ramón Hervás (UCLM, Computing Engineer)
  • Tania Mondéjar (UCLM, Psychologist)
  • Fco. Javier Navarro (UCLM, Geriatrician)
  • Luis Cabañero (UCLM, Computing Engineer)
  • Inocente Sánchez (UCLM, Computing Engineer)

Results

International Conference

Towards Abnormal Behaviour Detection on Elderly People at Home Through Smart Plugs and Its Relationship with Activities of Daily Living

by Adrián Sánchez, Jesús Fontecha, Iván González, Luis Cabañero , Christopher Nugent

International Conference on Ubiquitous Computing and Ambient Intelligence, Vol 835 118-123 (LNNS Springer) 2023. https://doi.org/10.1007/978-3-031-48306-6_12

This paper describes a proposal to monitor load measurements using smart-plugs connected to appliances distributed in a home environment, in order to detect specific activities of daily living. A recurrent neural network has been trained based on previously acquired time series, from such monitoring, framed at different time periods. The preliminary results of this work present how it would be possible to detect and associate some tasks at home with activities of daily living, generating activity patterns and, therefore being able to detect abnormal behaviours to act conveniently with the aim of increasing the life quality of elderly people at home.

https://link.springer.com/chapter/10.1007/978-3-031-48306-6_12

Final Master's work

Detection of anomalous behavior and their relationship with Activities of Daily Living

by Adrián Sánchez. Advisors: Jesús Fontecha, Iván González

Master's degree in computer engineering. Final work, presented and defended on July, 2024. <>

This Master Thesis aims to provide a solution to the detection of anomalous behaviors
that take place in the daily routines of an elderly person at home, through the monitoring and
analysis of data from connected devices distributed in the environment, as a facilitating element
for caregivers and professionals, in order to improve the quality of life of the person and preserve
their autonomy as long as possible. A system will be developed that will collect data and perform
continuous monitoring from devices connected to different appliances distributed in the home.

https://link.springer.com/chapter/10.1007/978-3-031-48306-6_12

Rust API

Library implemented in RUST to extract TAPO P110 smartplugs information and storage for ADLs in-depth analysis.

by Adrián Sánchez. Property of: MAmI lab

Rust library. Readme and code to extract and storage data from connected devices to TAPO P110 smartplugs.

Code in Rust to extract information from TAPO P110 smart plugs and store it in a MongoDB Atlas database. I will allow in-depth analysis about the correlation with Activities of Daily Living and its implication in the detection of abnormal user's patterns in order to improve his/her quality of life.

https://github.com/MAmILab/TAICare-api-extractionSP