Measuring vigilance level in demanding work setting

A.S. Smit¹, J. Benito², J. Troost² and C.H.M. Nieuwenhuis¹

¹Decis Lab, Thales Research & Technology, Delft, The Netherlands
²Thales Nederland, Hengelo, The Netherlands

Vigilance level strongly affects performance, especially on cognitively demanding tasks. Therefore, it is important to be able to determine the level of vigilance of individual people in many work settings. For example, an operator in a control room setting often needs to work in cognitively demanding situations under high mental work load. Traditionally, the level of vigilance has been determined using three different kinds of measures: physiological, performance and subjective. As they all have their own drawbacks, more stable measures can be obtained when they are simultaneously collected from different sources. It has proven very difficult to use physiological variables as indices of vigilance level on an individual basis. Many physiological parameters are very useful for distinguishing between experimental groups, but it appears rather difficult to determine a person’s vigilance on the basis of physiological data only. Moreover, EEG equipment can be rather intrusive in a work setting and EEG measurements are vulnerable to many artefacts. It has been demonstrated that vigilance can be estimated with much simpler measures such as subjective alertness scales and performance measures. In fact, on group level subjective vigilance questionnaires can be as reliable as physiological data when determining the level of vigilance. Subjective alertness can be assessed with a relatively small set of questions and performance measures can be acquired during tasks that are non-intrusive in the work environment and do not need to take much time to perform. We will describe a method for measuring vigilance based on the aggregation and fusion of individual measures. This method will not use physiological measures, but focuses on more easily applied behavioral measures, both subjective and objective, which can easily be derived in a work setting. The long-term goal is the development of a method that can be used in fully operational situations in the work setting, in order to enhance total system performance.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.

© 2005 Noldus Information Technology bv