PERFOMANCE MODELLING IN EO/IR SENSORS - Study of human perception thresholds in imaging systems

Financiador

Organização

Publicações

Development of a multimodal system for mental workload assessment

Publication . Mendonca, Pedro; Abreu, Manuel Adler

The main objective of this thesis was to develop a tool for Mental Workload (MWL) assessment, trying to answer to the growing interest of the community in evaluating the importance of this factor in task performance and human-machine interaction. The relevance of MWL in these areas has been the focus of an increasing number of scientific studies and publications despite the concept or definition of MWL in itself still fuzzy and the focus of many discussions. Furthermore, the objective assessment of this factor is still ill-defined, resourcing to many different methodologies and tools. Being a multivariable concept, Mental Workload should be assessed in a multimodal way. We used subjective methods including NASA task-load index; performance methods such as reaction time to make a decision; and objective methods including EEG device for brain related metrics, PPG for heart metrics and an eye-tracker for eye related metrics. In the case of the objective methods, the main issue for any sensor is the preprocessing stage, taking into account different type of noise and assuring a good quality data for reliable results. In order to contribute to the study of MWL and of its impact in task performance, we developed a tool based on the following assumptions: • Use of a non-medical-grade equipment in assessing mental workload induced by visual tasks; • Increase the MWL classification efficiency by correlation of different physiological measures, with different sensors; • Provide a common platform for sensor integration and signal processing, based on off-the-shelf, open source tools existing in the market. In this thesis, we showed how a multimodal system for MWL assessment can be made without resourcing to high cost / high complexity equipment, therefore enabling easier access for people interested in working in this field of study.

2023-06Tese de doutoramento

Acesso aberto

Ver mais

Entidade financiadora

Fundação para a Ciência e a Tecnologia

Programa de financiamento

FARH

Número da atribuição

PD/BD/52329/2013