Estimating the Driver’s Workload : Using Smartphone Data to Adapt In-Vehicle Information Systems (GLA)

Research Question: How can the driver’s workload be estimated in order to adapt information and entertainment systems?
Approach: Smartphone sensor data, situational factors and basic user characteristics are collected. This data is tested whether it significantly influences workload and can be used to estimate it.
Method: Workload is measured with a smartphone-based representation of the NASA-TLX and the RSME during a user study with 20 participants on different road types.
Results: Driving situation, gender and driving frequency significantly influence workload. Using only this information and smartphone sensor data the driver’s current workload can be estimated with 86% accuracy using a decision tree.

Keywords: Driver’s Workload, Workload Estimation, In-vehicle information systems.

Gerhard Lustig Prize Presentations
Location: Lecture Hall Aula Magna Date: May 20, 2015 Time: 12:30 pm - 12:50 pm christina_ohm_ma Christina Ohm