Dataset to predict mental workload based on physiological data



A high mental workload reduces the human performance and affects his/her ability to achieve a task. Despite the recent advances in neuroscience, yet there is a lack of knowledge about the interrelation between the mental processes in the brain and the produced mental workload at a giving time. The use of neuro-physiological data to assess abnormal mental states in the last decade has led to new manners to explore the brain and its association with mental workload. We present an open dataset for mental workload investigation. The dataset contains neuro-physiological recordings collected using an electroencephalogram (EEG) and an electrocardiogram (ECG). Participants were submitted to different tasks under different conditions to induce different levels of workload. In particular, three subsets were collected. First, playing the N-back test game to enforce the short term memory. Second, playing the Heat-the-Chair game (a serious game of own design) which enforce the performing of simultaneous tasks. Third, flying in an immersive simulated Airbus320 cockpit environment. The pilot must solve diverse critical situations, such as an engine failure, a sudden wind shear, or an urgent call of the air traffic controller (ATC). The design of the datasets has been validated by correlating the performance of subjects to their self-perceived difficulty. To make the dataset useful for testing the experiments, the ground-truth of mental workload of each task, both the objective and the subjective self-perceived is provided.
Date made available9 Jun 2022

Cite this