Researchers from the IT PIA-Lx group, in collaboration with Dutch computer vision specialists VicarVision and biomedical engineering spin-off PLUX, are currently conducting a large scale study to support the creation of novel tools for analysis of emotional wellbeing. The goal is to develop machine learning algorithms that can predict concerning changes in the health status, such as enhanced stress, based on indicators extracted from the interaction with everyday use devices (in particular, a tablet).
The experimental procedure involves the acquisition of video and physiological sensor data during the interaction between the participants and the tablet. Extracted modalities include facial expressions, respiration, Electrocardiography (ECG), and Electrodermal Activity (EDA). Participants attend two in-person sessions of 30-45 minutes, performed in consecutive weeks. In addition to the sensor data, clinically validated questionnaires are used to obtain ground truth data about the state of the user.
Anyone aged between 25-75 years old can participate and potentially contribute to advance the state-of-the-art in emotional wellbeing analysis. One of the envisioned outcomes of the work is a mobile app that can provide relevant information about the user health status in a pervasive approach, without requiring specialized setups.
Recruitment of volunteers is still ongoing at IT IST (Instituto Superior Técnico, Torre Norte - Piso 10), and the participation is rewarded with a €10 gift card once the two sessions are completed. If you’re interested in supporting this work or know anyone who might be, you can register using the online form available in the link bellow or by contacting the researcher conducting the study (firstname.lastname@example.org).
Aging aircrafts, transmission pipelines, bridges, railways, power plants and other large sections of our civilian and defense infrastructure are known to have far exceed their design lifetime. This has driven to an increasing demand of novel non-destructive evaluation (NDE) systems. NDE is an important research area that comprises the development of technologies and methodologies for assessing the structural integrity of critical materials, components and structures without affecting its future usefulness.
Gathering researchers from the Instrumentation and Measurements group, the Applied Electromagnetics group and the Optical Networking group, under the coordination of Helena Ramos, this internal IT project intended to increase accuracy and reliability to the inspection procedure by integrating complementary information obtained with two NDE techniques. The aim was to propose novel methods that push the state of art in NDE with a goal to monitor, detect and characterize hidden damage in carbon fiber polymer reinforced (CFRP) composites.
CFRP materials are increasingly being used in structures due to their lighter weight and improved strength, when compared to metals, in addition to their capability to be readily formed into custom shapes to meet unique requirements. However, NDE techniques have been problematic to employ in these new materials.
The RELIM research team joined two complementary NDE techniques: eddy current testing (ECT) and guided wave ultrasonic testing (GWUT). These two techniques were chosen, not only because they are complementary, but also because GWUT possesses propagation capabilities that enable inspection to be carried out with reduced scanning, allowing detection of damages in hidden parts of the structure or with difficult access. In addition, the use of tiny piezoelectric as sensors and actuators with GWUT is foreseen as a step forward in the direction of the development of a Structural Health Monitoring (SHM) evaluation. It is expected that in the near future, SHM shall bring huge costs and safety savings to different industries.
Photo: Locating real time impacts with PZT transducers in CFRP plates. After dropping the impactor through a tube, the guided wave signals obtained by the four PZTs are acquired by the digital oscilloscope.