● Developed multiprocess Python and C++ real-time software simulation mode, enabling three teams to prototype and satellite operators to rehearse procedures without relying on limited hardware in the loop testing resources

● Enhanced sim fidelity with power modeling, solar array deployments, hardware in the loop testing, and advanced analysis tools, enabling realistic behavior, improving developer productivity, and de-risked mission dangers

● Developed Python hardware & firmware emulators to enable pure software in the loop testing by mimicking UART communication between flight hardware and firmware, reducing the need for limited hardware resources

● Created sun sensor analysis/optimization tool, reducing coverage gap 5x & explaining post-launch eclipse anomaly

● Updated flight computer solar array and photodiode drivers for emergent hardware needs

● Supported satellite flying rehearsals, and went on console for post-launch activities with subsystem engineers

● Designed RSLQR feedback controller for fixed wing UAV, providing safety guarantees & better command tracking than previous PID controller

● Created state space autopilot model & simulation in Matlab/Simulink and verified performance in X-Plane flight simulator

● Developed Python simulation environment for double deep Q reinforcement learning for UAV path planning [1]

[1] Theile, M., Bayerlein, H., ​Nai, R​., Gesbert, D., & Caccamo, M. (2020). UAV Coverage Path Planning under Varying Power Constraints using Deep Reinforcement Learning. ​arXiv preprint arXiv:2003.02609​. [Accepted at IROS 2020]

● Prototyped high pin count testing mode to enable 6x greater product verification rate for Optane​®​ product by interfacing with B6700 series memory burn in tester and updating assembly like ALPG pattern test code

● Wrote, updated and cleaned C++ and Python infrastructure to manage tester data and removed unsafe code, massively reducing product verification overhead

● Enhanced compiler warnings and fixed existing issues with verification code for the next generation line of Optane​®​ chips, ensuring future changes to the codebase would be higher quality

● Programmed KAREL script to jog FANUC industrial robot, enabling demo for multi million € client

● Employed UDP socket messaging to communicate with Django Rest frontend to retrieve robot commands