Technische Universität München Robotics and Embedded Systems
 

Dipl.-Ing. Tobias Dirndorfer

 
E-Mail dirndorfer@ini.tum.de
Room MI 03.07.054
Phone +49.89.289.18102
Fax +49.89.289.18107
Address Institut für Informatik VI
Technische Universität München
Boltzmannstraße 3
85748 Garching bei München
Germany
Homepage
Dipl.-Ing. Tobias Dirndorfer
 

Curriculum Vitæ

2008 Ingolstadt Institute der Technischen Universität München (INI.TUM)
Research Assistant at the chair of Robotics and Embedded Systems under Prof. Knoll
Project "Development of predictive safety functions for the activation of restraint systems" in cooperation with AUDI AG
2008 Diploma in Mechanical Engineering, Technical University of Munich
2000 Abitur, Adalbert-Stifter-Gymnasium Passau
1980 Born in Passau, Germany

Project Group

Project


Development of predictive safety functions for the activation of restraint systems

Problem

The activation of restraint systems (airbags, seat belts) in accident situations is so far based on sensor signals (acceleration, pressure, impact sound), which have to be evaluated within a short time interval after the moment of collision. In addition to the limited time period, during which safety measures can be initiated, and the large number of necessary sensors this results in a strong dependence of implemented crash functions of the vehicle structure. The increasing performance of predictive vehicle sensors allows the development of new safety systems. These systems can support conventional crash sensors by preparing restraint systems in case of an approaching accident and activating them reliably and precisely timed depending on situational aspects.

Objective

The objective of this project is the development of a model-based functional concept, which chooses the best sensor information in terms of the current environmental conditions and uses it for the preparation and activation of restraint systems. The triggering of the restraint systems is supposed to be initiated on the base of an extensive recognition, analysis and predictive interpretation of the environmental situation.

Procedure

- Requirement definition and development of a concept based on an overview over the state of research and related projects
- Development of a method for the evaluation of object data (velocity, angle, offset, type, size, …) and object tracking under the requirements of accident scenarios
- Prototypic integration of the developed functional modules into the simulation as well as into a concept car and test of the overall functionality
- Implementation and integration of the functional modules into a serial control unit
- Evaluation of the developed functions

Outlook

Whereas the restraint systems are initially supposed to be activated after the time of collision but more robust and situation-specific than within common systems, it is the long term aim to already initiate the activation before the time of collision only by predictive sensors and car2x communication.

Predictive crash recognition

Predictive crash recognition


Publications