Böblingen, Baden-Württemberg, Germany
I am an open minded, enthusiastic and thoroughly positive data guy, being curious about various application domains for data science and machine learning. I love being creative and inventing things that did not exist before as well as working with inspiring people enriching discussions with their diverse cultural and academic backgrounds.
- Managing the development of various data products and complex analytics workflows with a mixed team (internal/external/off-shore) - Defining strategies for improving system performance in close collaboration with development teams of relevant comfort driving systems (TSA, ACC, LCA, ALSA)
- Developing data pipelines, analysis and dashboards about ADAS usage & performance in the customer fleet supporting the continuous improvement of ADAS features
- Working on my Phd thesis in the field of motion prediction for automated driving - Publishing and presenting academic papers in renowned international journals and professional conferences - Supporting the development team performing crowd analytics for automated driving
- Developing a system for the evaluation of road surface conditions and pothole detection using standard vehicle sensor data. - Presenting and publishing the work as conference article at VEHITS 2019 under the title 'CPD: Crowd-based Pothole Detection'.
- Assisting the researchers in the TAB-project "ROGER" (see: http://www.roger-projekt.de/) by doing independent research in the context of 360° camera systems, especially through selection, test and integration of 360° camera systems for pedestrian tracking and reidentification purposes.
- Developing a system for long-term vehicle position prediction in highway scenarios. Different regression methods were evaluated against each other and a classifier for the detection of lane-change maneuvers was designed using Python. - Documenting the whole work as Bachelor thesis.
- Supporting the development team in the task of risk estimation of lane-change events. Especially by implementing a GUI that allows the user to run scripts with dynamic parameters. - Implementing a procedure to calibrate risk estimates into probabilities, using Python.