CoachMyLife project STRAW project Insension project WellCo project
CrowdHealth project HeartMan project e-Gibalec project Fit4Work project
Commodity12 project Chiron project Confidence project PUBLICATIONS

I use the term ambient intelligence for all the research on the interpretation of sensor data, mobile and context-sensitive applications. This research area corresponds to (or maybe intersects with) the areas of ubiquitous and pervasive computing. I am particularly interested in analyzing human physical and psychological condition and behavior with sensors, but I occasionally also work on building automation and related topics. My main domain of interest is health, and because of that I throw in some health-related research that has little to do with ambient intelligence.

This page first lists all the projects I worked on (as a principal investigator, with the exception of the Confidence project), followed by publications. Quick links to these things can be found on top of the page.

A repository of ambient-intelligence datasets used in our research


CoachMyLife (2019-2022)

In this AAL project we are developing a system that will help seniors with memory impairment better tackle their activities of daily living. We will do this by recognizing their activities and needs using computer vision and other types of sensing, and providing reminders and instructions adapted to their current situation. Our department is focusing on computer vision and sensing with wearables.

STRAW (2018-2022)

The goal of this Flemish-Slovenian project is to improve the understanding of risk factors for day-to-day stress in the workplace. We are also developing methods for stress detection from physiological signals and behavioural cues obtained from smartphones.

Insension (2018-2021)

The goal of this H2020 project is to interpret the behaviour of people with profound intellectual and multiple disabilities, and to provide them with a rudimentary control over digital services by reacting to their expressions of pleasure and displeasure, as well as some other communication attempts. Our department is developing methods for extracting physiological signals from video, and for high-level interpretation of the user's behaviour.

Project video

Photoplethysmogram (PPG) reconstruction from video

WellCo (2017-2021)

This is a H2020 project developing a virtual coach to advise seniors on wellbeing and health. Our deaprtment is developing methods for nutrition monitoring with food-frequency questionnaires and sensors in a smartwatch. We are also working on emotion detection from speech, which is intended to improve spoken interaction with the coach.

CrowdHealth (2017-2020)

The goal of this H2020 project was to develop a platform to analyse big data related to health, in order to craft better public health policies. Our department focused on forecasting, clustering and risk assessment methods. We closely collaborated with the Faculty of Sport on the SloFit dataset, which is collected throught a national surveillance system for physical and motor development of children and youth.

HeartMan (2016-2019)

I was the coordinator of the HeartMan H2020 project, which developed a personal health system for congestive heart failure patients. The system consists of a mobile application, a sensing wristband, a backend for data management, and a web application for medical professionals. It provides comprehensive advice on the self-management of heart failure, ranging from physical exercise, nutrition, medication and self-monitoring, to psychological support. Our department adapted the methods for physical activity monitoring from the previous Fit4Work project, and developed methods for blood pressure estimation from photoplethysmogram, as well as the decision support system for providing the advice to the patients. The project was inspired by our previous work on the Chiron project.

Project video

Blood pressure estimation when stationary

Blood pressure estimation after exercise

e-Gibalec (2015)

This was a national project in which we developed a mobile application to encourage children to exercise. The application was based on gamification: the children could earn virtual currency by being physically active, and buy equipment for their avatars with this currency. The physical activity was detected automatically via smartphone sensors, or entered manually and confirmed by parents. The collected data was accessible to the users' physical education teachers.


Fit4Work (2014-2017)

This was an AAL project in which we developed a mobile application to help older workers improve and maintain their physical fitness, cope with mental stress and manage their work environment (temperature, humidity, CO2 level ...). The application was connected to a wristband with physiological sensors that was used for activity monitoring and stress detection. Our department adapted the methods for physical activity monitoring from the Commodity12 project to use the wristband, and developed methods for stress monitoring and recommending best actions to improve the work environment (e.g., opening the window).

Promotional video

Activity and stress monitoring

Commodity12 (2013-2015)

This was an FP7 project in which we developed a personal health system for diabetic patients. The system consisted of a mobile aplication and an optional ECG monitor (with an embedded accelerometer) for cardiovascular comorbidities. Our department developed methods for physical activity recognition and human energy expenditure estimation with the smartphone and/or the ECG monitor, which automatically adapted to the current device configuration. We also analysed ECG data to predict glycemias.

Human activity recognition and energy expenditure estimation

Chiron (2010-2014)

This was a large Artemis project that developed a framework for patient monitoring and decision support at home and in clinical environments. Our department developed methods for physical activity monitoring of congestive heart failure patients with wearable sensors, and a decision support system to assess their health risk.

Confidence (2008-2011)

This FP7 project was my first foray into the area of ambient intelligence. The goal of the project was to detect falls and other health problems of the elderly by recognizing anomalies in their posture and movement. This was done using location sensors that detected one to four tags worn on the user' bodies. Our department developed method for fall detection and the detection of other health problems.













2010 and earlier

Other documents