Oguz Altan

Title: Tracking and Evasion using Co-Training with Context Knowledge Optimizing unmanned aerial vehicle (UAV) trajectories for target tracking within urban environments - Employed reinforcement learning and deep learning with a game-theoretic co-training approach. - Utilized procedural map generation technique for generating maps resembling typical urban layouts. - Integrated convolutional neural networks (CNNs) into reinforcement learning models to process map images and enrich agent observations.

Worked in the Technology Department of Siemens focusing on the additive manufacturing project called Wire Arc Additive Manufacturing (WAAM). Responsible for designing the artificial intelligence system and integrating this new method. - Anomaly detection using Python-based machine learning algorithms. - Process monitoring. - Automation software development. - Edge computing with NVIDIA Jetson. - Closed-loop controller optimization.

Worked in the Department of Information Technology and Communication Electronics, collaborating with the Embedded Microcontroller Systems Laboratory and IoT teams, a joint effort of the Fraunhofer Institute and FAU Erlangen-Nürnberg. - Designed Printed Circuit Boards (PCB) and programmed microcontrollers for both teams. - Collaborated with academics, including Ph.D. students, postdoctoral researchers, scientists, and engineers.

Payload Electronics Design team - Designed an Instrumentation Amplifier (OPAMP), employing differential signaling for noise reduction and allowing dynamic gain adjustments to modulate signal amplitude. - Converted the analog signal to digital using the MCP3204 12-bit 4-channel analog to digital converter (ADC). - Facilitated the display of the digital representation of the analog input voltage by the integration with an Arduino via the Serial Peripheral Interface (SPI) communication protocol.

Major: Systems and Automation