Neues Forschungsprojekt zu Autonomem Fahren: Nutzung und Analyse von Crowd- und Kontext-Daten

Universität Ulm

Das Institut für Datenbanken und Informationssysteme (DBIS) setzt seine langjährige Zusammenarbeit mit der Daimler AG fort. Im gemeinsamen Forschungsprojekt PRoTECT  (PRedicting The driving behavior under varying Environmental ConsTraints) arbeitet der DBIS-Doktorand Florian Wirthmüller im Bereich des autonomen Fahrens an fortschrittlichen Algorithmen zur Prädiktion des Fahrverhaltens unter Nutzung von Crowd- und Kontext-Daten.


Hintergrund (in Englisch)

To enable modern cars to drive autonomously or to assist drivers properly in various scenarios and environments are some of the most challenging undertakings of our time. To realize this vision, cars need to analyze and understand their environment as well as the current traffic situation in real-time with high accuracy. In this context, prediction methods are required to enable the planning of a secure driving corridor. Though current Advanced Driver Assistance Systems (ADAS) already enable such predictions, most of them only apply simple heuristics, utilizing the data directly provided by the sensor systems. The PRoTECT project, therefore, aims to improve prediction accuracies by additionally incorporating contextual information, i.e., information a human driver would conclude from the context of the situation and his knowledge thereof in order to adapt his behavior, e.g., to drive more carefully in the case of a slippery road. In addition, location-based particularities (e.g. crowded roads like in India or rectangular roads like in Manhattan) will be taken into account to broaden the use of driver assistance systems.