Mohamed Sabry Moustafa

M.Sc. Thesis Student

Robert Bosch GMBH•  November 2022 - April 2023

M.Sc. Thesis student in the Deep Learning Loop for Autonomous Driving Department.

Universität Stuttgart

Information Technology, MS•  November 2020 - Present

Master's in the field of Information Technology (Major: Computer Software & Hardware Architecture). I am specialized in Deep Learning, NLP, and Computer vision.

Technische Universität München

Electrical Engineering & Computer Science, BS•  March 2019 - August 2019•  GPA: 1.3

My thesis dealt with incorporating a feasible and efficient legibility structures into autonomous highway driving scenarios. The proposed method is elaborated with respect to a merging highway scenario and a full highway scenario incorporating multiple autonomous vehicles interacting with each other to allow a specific vehicle to merge safely and efficiently while each other vehicle is performing a specific maneuver to allow for the merging to happen successfully. Therefore, legibility cost functions and inference functions, which formulate the perceived probability of motion of each vehicle, are elaborated. These structures are used by the Model Predictive Control solver in order to optimize the trajectory of each vehicle such that the resulting trajectory of each vehicle is legible and understandable for other vehicles in the same environment. Moreover the assumed methods are evaluated with respect to their cost efficiency for the system of vehicles.