Using ontologies for dataset engineering in automotive AI applications
Martin Herrmann1,a, Christian Witt2, Laureen Lake1,b, Stefani Guneshka3,d, Christian Heinzemann1,c, Frank Bonarens4, Patrick Feifel4 and Simon Funke3,e
1Robert Bosch GmbH
aMartin.Herrmann@de.bosch.com
bLaureen.Lake@de.bosch.com
cChristian Heinzemann@de.bosch.com
2Valeo Schalter und Sensoren GmbH
christian.witt@valeo.com
3Understand AI
dstefani@understand.ai
esimon.funke@understand.ai
4Stellantis, Opel Automobile GmbH
frank.bonarens@stellantis.com
ABSTRACT
Basis of a robust safety strategy for an automated driving function based on neural networks is a detailed description of its input domain, i.e. a description of the environment, in which the function is used. This is required to describe its functional system boundaries and to perform a comprehensive safety analysis. Moreover, it allows to tailor datasets specifically designed for safety related validation tests. Ontologies fulfill the task to gather expert knowledge and model information to enable computer aided processing, while using a notion understandable for humans. In this contribution, we propose a methodology for domain analysis to build up an ontology for perception of autonomous vehicles including characteristic features that become important when dealing with neural networks. Additionally, the method is demonstrated by the creation of a synthetic test dataset for an Euro NCAP-like use case.
Keywords: ontology, dataset engineering, autonomous driving, artificial intelligence, neural network.
