Research Article
Constraint-based Context-Rule Representation and Risk Classification for Driver Assistance Systems
@INPROCEEDINGS{10.4108/ICST.ISVCS2008.3602, author={Simone Fuchs and Stefan Rass and Kyandoghere Kyamakya}, title={Constraint-based Context-Rule Representation and Risk Classification for Driver Assistance Systems}, proceedings={1st International ICST Symposium on Vehicular Computing Systems}, proceedings_a={ISVCS}, year={2010}, month={5}, keywords={Context-awareness Constraint Programming Fuzzy-probabilistic Risk Classification Overtaking Assistance}, doi={10.4108/ICST.ISVCS2008.3602} }
- Simone Fuchs
Stefan Rass
Kyandoghere Kyamakya
Year: 2010
Constraint-based Context-Rule Representation and Risk Classification for Driver Assistance Systems
ISVCS
ICST
DOI: 10.4108/ICST.ISVCS2008.3602
Abstract
Driving context, including oncoming and approaching vehicles, traffic signs, road conditions and legal traffic law, has major influence on the recommended behavior of a driver. For a driver assistance system these legal and environmental constraints must be available to produce useful behavior recommendations. Using a contextaware overtaking assistant as demonstrational, conceptual reasoning component we present an approach for representation of legal and environmental context rules, exploiting constraint-logic programming, where spatio-temporal influence factors are modeled as dynamic constraints. A-priory fuzzy risk-classification is done to point out potential risks. The probability of an accident is then obtained for the recommended maneuver, using information from the driver context.