Methodological aspects of naturalistic driving studies and field operational tests
Partner
FAT (Forschungsvereinigung Automobiltechnik e.V.)
Description
In recent years, there have been conducted several so called naturalistic driving studies, which do not follow an experimental approach, but try to describe natural driving behaviour, based on mass data, in a descriptive way.
In most cases, such studies try to evaluate the impact of new systems/measures on driving behaviour, sometimes also expressing support needs in everyday driving. The typical procedure is the following:
- equip multiple experimental vehicles with measuring instrumentation
- give them to a large number of participants for uninstructed, "natural" everyday use
- collect data
- analyse the data
- interpret the results
One major drawback of such an approach is a certain lack of objectivity in terms of independence from the experimenter. On the other hand, the approach represents a classical way for describing initial situations, identifying references or testing new systems.
This project aims at developing methodological and technical guidelines and tries to include the following aspects:
- reduction of planning/technical effort
- description and discussion of technical attemps for data acquisition
- discussion of methods to increase the significance of driving data that usually is of high variability and low standardisation
- discussion of contrary results of experiment vs. field studies using available data
- deduction of recommendations for samples (number of drivers, vehicles, ...), data to log, data reduction, statistical analysis
- definition of research needs
Contact
People
Dipl.-Psych. Matthias Henning, Dipl.-Ing. Eric Richter, Dipl.-Inf. Holger Lietz, Prof. Josef Krems, Prof. Gerd Wanielik
