Pratik Nadkarni

Master thesis with title : Investigating the applicability and impact of an event-based temporal contrast vision sensor in the automotive context. Description : This thesis aimed at investigating the possibilities and scope of using a novel asynchronous eventbased vision sensor in automotive applications currently facing limitations with a conventional synchronous frame-based image sensor.

-- Lane change prediction : Set-up & integration of the driver face tracking as a driver model input in the case study for developing a lane change prediction model based on driver’s head rotation. -- Driver Fatigue : Developed a conceptual rule-based model for driver fatigue detection based on non-intrusive technique for comparing rule-based & machine learning-based approaches.

-- Driver Model : Implemented a real-time face tracking API within the ROS framework in the BMW prototype. Conducted post-processing of the simulator test data for developing the driver model to predict driver vigilance based on gaze direction. -- Environment Model : Developed a GUI toolchain in MATLAB for analysing the detected pedestrian movement behaviour patterns for developing a classification model based on situation criticality.

-- Conducted endurance & performance testing of power train & hydraulic components of a tractor. -- Supported the design & implementation of test cycles for life-cycle testing. -- Contributed in design & development of new test beds.

Active safety : Driver assistance systems + Sensor technology & signal processing Power train : Internal combustion engines + Applications of engine control units

Electrical drives, robotics, Pneumatics-hydraulics, Automotive mechatronics.