Wednesday 18 February 2009

Andreas Riener defends his PhD in Linz

After a stop-over in Stansted/Cambridge at the TEI conference I was today in Linz, Austria, as external for the PhD defense of Andreas Riener. He did his PhD with Alois Ferscha and worked on implicit interaction in the car. The set and size of experiments he did is impressive and he has two central results. (1) using tactile output in the car can really improve the car to driver communication and reduce reaction time. And (2) by sensing the force pattern a body creates on the seat driving relates activities can be detected and to some extend driver identification can be performed. For more details it makes sense to have a look into the thesis ;-) If you mail Andreas he will probably sent you the PDF...

One of the basic assumptions of the work was to use implicit interaction (on input and output) to lower the cognitive load while driving - which is defiantly a valid approach. Recently however we also discussed more the issues that arise when the cognitive load for drivers is to low (e.g. due to assistive systems in the car such as ACC and lane keeping assistance). There is an interesting phenomenon, the Yerkes-Dobson Law (see [1]), that provides the foundation for this. Basically as the car provides more sophisticated functionality and requires less attention of the user the risk increase as the basic activation of the driver is lower. Here I think looking into multimodality to activate the user more quickly in situations where the driver is required to take over responsibility could be interesting - perhaps we find a student interested in this topic.

[1] http://en.wikipedia.org/wiki/Yerkes-Dodson_law (there is a link to the 1908 publication by Yerkes, & Dodson)