Johannes Bach studierte Elektro- und Informationstechnik am Karlsruher Institut für Technologie (KIT). Schwerpunkte während seines Studiums waren Industrielle Informationssysteme, Signalverarbeitung und Regelungstechnik. Seine Diplomarbeit mit dem Titel "Development of a GNSS Interference Detection and Characterization Toolbox" schrieb er bei EADS Astrium in Ottobrunn. Seit November 2011 ist er als wissenschaftlicher Mitarbeiter der Forschungsabteilung ESS am FZI tätig.
Aktuell beschäftigt er sich mit der Entwicklung von Methoden und Werkzeugen zur Entwicklungs- und Testunterstützung.
- Entwicklungsmethodik Fahrerassistenzsysteme
- Test- und Validierung von Fahrzeugfunktionen
- Systemidentifikation und -modellierung
- Reactive-Replay Approach for Verification and Validation of Closed-Loop Control Systems in Early DevelopmentInfoDetails
Johannes Bach, Marc Holzäpfel, Stefan Otten, Eric Sax, 2017
Enhanced technological capabilities render the application of various, increasingly complex, functional concepts for automated driving pos-sible. In the process, the significance of automotive software for a satisfactory driving experience is growing. To benefit from these new opportunities, thorough assessment in early development stages is highly important. It enables manufacturers to focus resources on the most promising concepts. For early assessment, a common approach is to set up vehicles with additional prototyping hardware and per-form real world testing. While this approach is essential to assess the look-and-feel of newly developed concepts, its drawbacks are re-duced reproducibility and high expenses to achieve a sufficient and balanced sample. To overcome these drawbacks, new flexible, realis-tic and preferably automated virtual test methods to complement real world verification and validation are especially required during early development phases. In this contribution, we present a method for au-tomated system assessments based on the reuse of recorded driving data in closed-loop simulation and its application in early develop-ment of a predictive cruise control system. Firstly, we identify the re-quirements for early assessment of closed-loop system concepts, ana-lyze the eligibility of established methods regarding the identified re-quirements and describe open challenges for development of automo-tive software systems with focus on early development stages. Our previously introduced Reactive-Replay approach addresses these challenges by enabling reuse of recorded driving data in closed-loop simulation. We complement this approach by introducing automated assessments for evaluation of software increments. By integrating pe-riodic assessments into the development process, we achieve continu-ous tracking of software quality with very small effort. It is shown that the provision of a broad data pool for simulation based evalua-tion of new and refined concepts contributes to a substantial reduc-tion of real world test mileage in early development stages.
- Model based scenario specification for development and test of automated driving functionsInfoDetails
Johannes Bach, Stefan Otten, Eric Sax, 2016
Research and evaluation of algorithms and system architectures for automated driving advanced to a stage where transition from prototyping to series development seems practicable in some extent. While particular systems for environmental perception play a key role in advanced driving assistance systems, we still lack feasible methods for specification and validation of complex driving scenarios. This leads to increased effort in testing and inconsistent requirement definition along different development phases. In this paper we propose a methodology for abstract positional and temporal description of driving scenarios. The approach utilizes a movie related and omniscient view composed of sequential acts. Each act combines both states and interactions of distinct participants as well as the rudimental scenery. Selective events trigger changes in conduct leading to transitions between acts. Graphical visualization provides simple presentation of complex scenarios. Rule sets provide consistency checks and support semi-automated generation of test cases. The presented methodology facilitates model based test specification and requirements design constituting a consistent characterization of system environment from early concept and development to validation.
- Control based driving assisant functions' test using recorded in field dataInfoDetails
Johannes Bach, Kai-Lukas Bauer, Marc Holzäpfel, Martin Hillenbrand, Eric Sax, 2015
For verification of Advanced Driver Assistant Systems (ADAS) versatile types and levels of laboratory, as well as vehicular tests are applied. This paper presents a methodology for reutilization of recorded test data during virtual verification of control based ADAS. By changing the recorded data’s domain, coherent and reactive stimulus of the System Under Test (SUT) is enabled. To do so, identification of causal dependencies between various input data streams is required. The basic feedback loop of the control system is closed by means of a conventional plant model and all additional input data is derived from conditioned data records. Therefore, introducing reproducibility to real world Rapid Control Prototyping (RCP) test, omitting time demanding scenario parametrization and ensuring realistic scenarios. Our concept is explained and demonstrated using recorded data during development of a Predictive Cruise Control (PCC) system.
Export Suchergebnis .bib
Telefon: +49 721 9654-168