Emran Yasser Moustafa

My name is Emran Yasser Moustafa, and I am a first-year PhD student at the University of Edinburgh.

My PhD research focuses on developing shared autonomy for assistive driving applications. I’m particularly interested in automated decision-making and reasoning under incomplete information. My research is affiliated with the Centre for AI in Assistive Autonomy and I’m supervised by Prof. Subramanian Ramamoorthy.

Prior to my doctoral studies, I received First Class Honours in my BEng and MEng Electronic & Computer Engineering degrees at Trinity College Dublin. I completed my Master’s thesis with Prof. Ivana Dusparic, for which I received the Collen Award for the best thesis in Computer Engineering. I completed my Bachelor’s thesis with Prof. Julian Kooij during a study-abroad year at TU Delft. I was fortunate to receive both the Laidlaw and Kinsella scholarships, which funded my studies throughout university. I graduated with an honorary Gold Medal for academic excellence at Trinity College Dublin.

Outside of studying, I was a long-time member of Formula Trinity, where I had the opportunity to take part in F1Tenth (2021) and Formula Student UK (2022, 2023). During my time in Formula Trinity, I served as state-estimation lead and later as captain of the driverless team. Over the years, our team received multiple awards, and I was nominated by my teammates for the Craig Dawson Award (MVP) in 2022. Beyond Formula Student, I completed several research internships at Trinity College Dublin, FlyingLabs Madrid, and TU Delft. I also founded my own robotics community-building non-profit, RacingAI, which hosted robotics events across Ireland for a year and a half.

On a personal note, I enjoy food and drink, art, music, and Brazilian jiu-jitsu. I also love board games and am currently building out a small indie board game studio based in Dublin (currently in stealth mode). Feel free to get in touch! 🎯

news

Jul 16, 2026	I attended the pro² Device Prototyping Summer School at Lancaster University
Feb 10, 2026	I attended UK PlanSIG 2026 in Edinburgh. Thank you to all the speakers!
Oct 23, 2025	I have officially graduated from Trinity College Dublin
Sep 20, 2025	The paper I wrote from my Master’s thesis was accepted to IEEE ICAR 2025!
Sep 08, 2025	I have started my 4-year PhD at the University of Edinburgh

selected publications

Context-Aware Model-Based Reinforcement Learning for Autonomous Racing

Emran Yasser Moustafa and Ivana Dusparic

In IEEE International Conference on Advanced Robotics (ICAR), 2025

Abs

Autonomous vehicles have shown promising potential to be a groundbreaking technology for improving the safety of road users. For these vehicles, as well as many other safety-critical robotic technologies, to be deployed in real-world applications, we require algorithms that can generalize well to unseen scenarios and data. Model-based reinforcement learning algorithms (MBRL) have demonstrated state-of-the-art performance and data efficiency across a diverse set of domains. However, these algorithms have also shown susceptibility to changes in the environment and its transition dynamics. In this work, we explore the performance and generalization capabilities of MBRL algorithms for autonomous driving, specifically in the simulated autonomous racing environment, Roboracer (formerly F1Tenth). We frame the head-to-head racing task as a learning problem using contextual Markov decision processes and parameterize the driving behavior of the adversaries using the context of the episode, thereby also parameterizing the transition and reward dynamics. We benchmark the behavior of MBRL algorithms in this environment and propose a novel context-aware extension of the existing literature, cMask. We demonstrate that context-aware MBRL algorithms generalize better to out-of-distribution adversary behaviors relative to context-free approaches. We also demonstrate that cMask displays strong generalization capabilities, as well as further performance improvement relative to other context-aware MBRL approaches when racing against adversaries with in-distribution behaviors.
Road User Specific Trajectory Prediction in Mixed Traffic Using Map Data

Hidde JH Boekema, Emran Yasser Moustafa, Julian FP Kooij, and 1 more author

IEEE Robotics and Automation Letters, 2025

Abs

This paper studies road user trajectory prediction in mixed traffic, i.e. where vehicles and Vulnerable Road Users (VRUs, i.e. pedestrians, cyclists and other riders) closely share a common road space. We investigate if typical prediction components (scene graph representation, scene encoding, waypoint prediction, motion dynamics) should be specific to each road user class. Using the recent VRU-heavy View-of-Delft Prediction (VoD-P) dataset, we study several directions to improve the performance of the state-of-the-art map-based prediction models (PGP, TNT) in urban settings. First, we consider the use of class-specific map representations. Second, we investigate if the weights of different components of the model should be shared or separated by class. Finally, we augment VoD-P training data with automatically extracted trajectories from the 360-degree LiDAR scans by the recording vehicle. This data is made publicly available. We find that pre-training the model on auto-labels and making it class-specific leads to a reduction of up to 22.2%, 20.0%, and 18.2% in minADE (K=10 samples) for pedestrians, cyclists, and vehicles, respectively.