I am part of the vehicle buildup team, and responsible for the NVIDIA DRIVE AV software deployment. My main task is to make sure that the functional software stack consisting of perception, mapping and planning modules work as intended for vehicle automation. For this, I collaborate closely with other engineering teams.

Worked on a project for the development of a SAE L4 urban automated driving system. Defined the first release of the security test concept based on the complete system and software architecture with focus on risk mitigation against vehicle attacks. Analyzed threats against sensors and AV stack components, and developed test plans. Implemented security tests (e.g. QNX hardening), software test methods (e.g. AFL fuzzing), test automation on target hardware in CI fashion. Performed vehicle tests in the context of sensor security for radar and camera perception, as well as localization, and analyzed functional performance issues during attacks. Researched adversarial sensor attack potentials on deep learning models, and validated the system robustness during attacks in simulations and test track measurements. Developed a sensor security simulation environment based on the Unreal Engine with the Bosch Perception stack in open-loop to identify system vulnerabilities early on.

Worked on adaptive cruise control systems with radar and camera data fusion in the field of driver assistance systems. Developed and maintained radar software components related to digital signal processing, vehicle dynamics estimation, object classification and situational analysis. Gained firm knowledge in the analysis of vehicle measurements with focus on identification of target object loss reasons and object type misclassifications.

Worked on a project for the development of a novel automotive lidar sensor. Developed road lane and border extraction algorithms in C++. Implemented a test environment to analyze the detection performance using OpenCV. Experimented with the design and implementation of signal processing algorithms.

Worked on a project for the development of a 48V electrical boost recuperation machine. Implemented software requirements in Matlab/Simulink under consideration of the V-model specification and ISO-26262 compliance. Verified algorithms in Simulink and Hardware-in-the-Loop simulations. Performed tests by analyzing code coverage, runtime and controls.

Conducted a business case study in the field of rapid prototyping for flexible engine control. Designed a gasoline engine control interface implemented in Matlab/Simulink to set and monitor engine speed or load through fuel injection and ignition based on driver's demand. Tested the developed environment on target sensors and actuators at the customer site.