The duration of incidents in dynamic human-machine system represent crucial information for system analysis and the regulation of task-performance. With rising task-demands people tend to be less reliable in time-estimations and cause errors in their interaction with a complex system.
A model of time perception was developed in the course of this project, which posits that working memory mechanisms distort time perception under certain circumstances. This approach was integrated into a cognitive architecture. With such an integrated approach of human time perception it is possible to predict temporal human errors in the context of complex dynamical systems. In early phases of system development a system could be evaluated and blueprints could be favoured or sources of error could be detected or prevented.
In order to verify the developed quantitative model, predictions for variations of a counting task were generated. These predictions were compared with date of a series of three experiments. In the first experiment the connection between the numbers of presented lists on time estimations was evaluated. The model as well as the experimental data shows no such connection for prospective time estimation. Operators in process plants often experience severe changes in task demands and have to repeatedly estimate the duration of a certain processes. These situations are explored in the second experiment. A comparison with the model data reveals that predictions are met for the low to high demand switch. In the other condition some artefacts occurred that have to be explored in more detail. A third experiment was conducted with the simulation of a process plant. The experimental design was slightly modified due to the findings derived in the previous experiment. The experimental data of the third experiment corresponds to the model predictions.
The results show that the size, the spreading and the direction of time estimations are correctly predicted by the model. An additional influencing fact or on time estimation was found that occurs whenever participants experienced unexpected changes. This effect can also be explained by the theoretical account of time estimation but is difficult to model due to the absence of a modelling approach that explains how expectations develop during a task. The TaSTE (Task Sensitive Time Estimation) module is able to simulate most of the examined effects of human time-perception and therefore offers a tool which can be used for the evaluation of temporal aspects in the context of human-machine systems.
Team Time Cognition
Nele Rußwinkel 
|Title of Book||Fahrermodellierung in Wissenschaft und Wirtschaft.- 4. Berliner Fachtagung Fahrermodellierung, Berlin, 13./14. Juni 2013|
|Address||Düsseldorf: VDI Verlag|
|Editor||Jürgensohn, T. and Kohlrep, H.|