Search

The future generation user system interfaces need to be user-centric which goes beyond user-friendly and includes understanding and anticipating user intentions. We introduce the concept of operator models, their role in implementing user-system symbiosis, and the usefulness of brain-based indices on for instance effort, vigilance, workload and engagement to continuously update the operator model. Currently, the best understood parameters in the operator model are vigilance and workload. An overview of the currently employed brain-based indices showed that indices for the lower workload levels (often based on power in the alpha and theta band of the EEG) are quite reliable, but good indices for the higher workload spectrum are still missing. We argue that this is due to the complex situation when performance stays optimal despite increasing task demands because the operator invests more effort. We introduce a model based on perceptual control theory that provides insight into what happens in this situations and how this affects physiological and brain-based indices.We argue that a symbiotic system only needs to intervene directly in situations of under and overload, but not in a high workload situation. Here, the system must leave the option to adapt on a short notice exclusively to the operator. The system should lower task demands only in the long run to reduce the risk of fatigue or long recovery times. We end by indicating future operator model parameters that can be reflected by brain-based indices.

The Dutch Ministry of Transport, Public Works and Water Management needs knowledge in the area of workload in car driving. This paper describes a study in which the effects of traffic volume, road lighting, and a cognitively loading secondary task on driving behaviour and workload were investigated. In a field experiment, subjects drove a fixed route in a TNO instrumented vehicle on a motorway, on which the public lighting was experimentally manipulated. Measurements were started at the beginning of the evening, yielding runs in busy as well as in quiet traffic conditions. In half of the runs, subjects also had to perform a cognitively loading secondary task. During the runs we registered the driving behaviour as well as a number of physiological variables (ECG and EOG). After each run, subjects completed a questionnaire that measured the subjective rating of mental effort. We expected the workload to be higher in dense traffic, on an unlit road, and while executing the secondary task. Results showed that traffic flow intensity and the presence of the secondary task both affected speed choice and workload (both physiological and subjective measures). When conducting the secondary task, subjects carried out fewer overtaking manoeuvres. Motorway lighting especially affected the steering reversal rate, the eye blink frequency, and the steering effort.

The sensitivity of physiological measures to mental workload has been investigated in a flight simulator. Twelve pilots had to fly through a tunnel with different levels of difficulty. Additionally, they had to perform a memory task with four levels of difficulty. The easiest memory task was combined with the easiest tunnel task and the most difficult memory task with the most difficult tunnel task. Between the tunnel tasks, subjects had to fly a pursuit task in which a target jet had to be followed. Rest periods before and after the experiment were used as a baseline for the physiolog-ical measures. Mental workload was measured with heart period, continuous blood pressure, respiration and eye blinks. Heart rate variability, blood pressure variability and the gain between systolic blood pressure and heart period (modulus) was derived from the measures. All measures showed differences between rest and flight, and between the pursuit and the tunnel task. Only heart period was sensitive to the different levels in the tunnel task. Since heart rate variability was affected by respiration, it could only be interpreted together with the respiratory data. The modulus, how-ever, was hardly influenced by respiration and therefore, appears te be a better measure than heart rate variability. From the respiratory parameters, the duration of a respiratory cycle was the most sensitive to changes in workload. The time in between two successive eye blinks (blink interval) increased and the blink duration decreased as more visual information had to be processed. Increas-ing the difficulty of the memory task leaded to a decrement in blink interval, probably caused by subvocal activity during the rehearsal of the letters. Blink duration was not influenced by memory load.

Hartslag, bloeddruk en oogknippers geven goed verschillende niveaus van mentale inspanning weer. Hartslagvariabiliteit kan tijdens het uitvoeren van complexe taken verstoord worden door ademhaling.

The aim of the present study is to validate different measures for mental workload. Ten aspirant fighter jet pilots flew several scenarios in a flight simulator. The scenarios were divided into segments with different levels of task load. During the flight, heart rate, respiration and blood pressure were recorded. After the tasks, subjects rated the work-load for segments of the scenario with the NASA-TLX and the BSMI (a Dutch effort scale). Furthermore, cortisol was measured on the training and experimental day (before and after each scenario). Heart rate, heart rate variability and several respiratory parameters reflected the changes in task load. Blood pressure was not systematically affected by task load. The BSMI was more sen-sitive than the TLX; the task load effect was stronger and the correlations with the physiological parameters were higher. Cortisol levels were slightly increased on the experimental day but the post task levels did not differ from the pre-task levels. There was a correlation between mean cortisol levels and performance: subjects with high cortisol levels performed poorly.