Driver workload classification through neural network modeling using physiological indicators

Conference Paper (2013)
Author(s)

R.G. Hoogendoorn (TU Delft - Transport and Planning)

B van Arem (TU Delft - Transport and Planning)

Department
Transport and Planning
DOI related publication
https://doi.org/10.1109/ITSC.2013.6728565
More Info
expand_more
Publication Year
2013
Language
English
Department
Transport and Planning
Pages (from-to)
2268-2273
ISBN (print)
9781479929146

Abstract

Advanced Driver Assistance Systems may have a positive effect on traffic flow efficiency, the environment, safety and comfort. However these systems may have a negative impact on driving behavior following a change in driver workload. It is therefore crucial to develop a so-called driver workload manager. In order to manage driver workload an adequate classification of driver workload is indispensible. In this contribution we propose to classify and predict driver workload through physiological indicators of driver workload, driver characteristics and characteristics of the driving condition using a neural network modeling approach. We show that the proposed network yields a very good classification of driver workload. The contribution finishes with a discussion section and recommendations for future research.

No files available

Metadata only record. There are no files for this record.