Research field

Geometric and Riemannian robot learning

Learning and control are formulated directly on manifolds and Lie groups, not on flattened approximations of them. This keeps what a robot learns consistent with the structure of the space it moves in.

A sphere with a geodesic between two points, a tangent plane at the start and tangent vectors carried along the path.

About this field

Orientations, stiffness matrices and robot configurations are not flat vectors — they carry curvature and constraints. We formulate learning and control directly on these manifolds, so that generalization respects the structure of the space instead of working around it.

People

  • Hozefa Jesawada

    Hozefa Jesawada

    Postdoctoral researcher since 2025

    The research covers safe learning-based control for soft and continuum robots. It also develops Grassmannian methods for null-space control.

    Hozefa Jesawada is a Postdoctoral Researcher at New York University Abu Dhabi. He earned his PhD in Information Technologies for Engineering from the University of Sannio, Italy, and subsequently worked as a postdoctoral researcher at the University of Salerno. His research focuses on reliable decision-making and control for robotics, combining tools from optimization, uncertainty-aware learning, inverse optimal control, and robust planning.

    He is particularly interested in how principled learning methods can support embodied intelligence in real robotic systems, including manipulation, model-based formal control, and safety-aware autonomy.

  • Ikhlas Ben Hmida

    Ikhlas Ben Hmida

    PhD student since 2024

    The research covers the modelling and design of a soft prosthetic hand. It also explores data-driven decoding of hand and finger motion from surface electromyography.

    Ikhlas Ben Hmida is a Global Ph.D. fellow from NYU Tandon School of Engineering and a visiting graduate research assistant at the REACH Lab at NYUAD, where she is conducting her PhD research.

    Her research interests include soft robotics, bio-inspired systems, and medical applications of robotics. She is currently exploring data-driven approaches to decoding hand and finger motion from surface electromyography (sEMG) to advance human-machine interfaces and dexterous prosthetic and assistive hands.

  • Omar Mostafa

    Omar Mostafa

    PhD student since 2026

    The research is on stable robot learning, including stable dynamical-systems control. It also covers vision-language-guided policy learning and simulation environments for manipulation.

    Omar Mostafa is a Global PhD student in Electrical and Computer Engineering. His research interests are robot learning and control, with a focus on physics-informed reinforcement learning and world models for robotic manipulation.

    Previously: Graduated magna cum laude from NYU Abu Dhabi with a bachelor’s degree in Electrical Engineering.

  • Narayana Reddy Marthala

    PhD student since 2026

    The research covers contact control for space robots. The robots studied are dual-arm and free-floating.

Publications since 2024

  1. Constrained Trajectory Optimization on Matrix Lie Groups via Lie-Algebraic Differential Dynamic Programming

    Systems & Control Letters (SCL), vol. 204

    G. Alcan, F. J. Abu-Dakka, V. Kyrki

  2. MeshDMP: Motion Planning on Discrete Manifolds using Dynamic Movement Primitives

    IEEE International Conference on Robotics and Automation (ICRA)

    M. Dalle Vedove, F. J. Abu-Dakka, L. Palopoli, D. Fontanelli, M. Saveriano

  3. Learning Periodic Skills for Robotic Manipulation: Insights on Orientation and Impedance

    Robotics and Autonomous Systems (RAS)

    F. Abu-Dakka, M. Saveriano, L. Peternel

  4. EEG Motor Imagery Classification: Tangent Space with Gate-Generated Weight Classifier

    Biomimetics

    S. Omari, A. Omari, F. Abu-Dakka, M. Abderrahim

  5. A Unified Formulation of Geometry-Aware Discrete Dynamic Movement Primitives

    Neurocomputing

    F. J. Abu-Dakka, M. Saveriano, V. Kyrki

  6. Safe Execution of Learned Orientation Skills with Conic Control Barrier Functions

    IEEE International Conference on Robotics and Automation (ICRA)

    Z. Shen, M. Saveriano, F. J. Abu-Dakka, S. Haddadin

  7. CITR: A Coordinate-Invariant Task Representation for Robotic Manipulation

    IEEE International Conference on Robotics and Automation (ICRA)

    P. So, R. I. Cabral Muchacho, R. J. Kirschner, A. Swikir, L. Figueredo, F. J. Abu-Dakka, S. Haddadin

  8. G3VLA: Geometric inductive bias for Vision-Language-Action Models

    Preprint arXiv preprint — not yet peer-reviewed

    Yue Peng, Yongzhe Zhao, Artur Habuda, Khuyen Pham, Yanheng Zhu, Tran Nguyen Le, Fares Abu-Dakka, Li Guo

  9. Stable Robot Motions on Manifolds: Learning Lyapunov-Constrained Neural Manifold ODEs

    Preprint arXiv preprint — not yet peer-reviewed

    David Boetius, Abdelrahman Abdelnaby, Ashok Kumar, Stefan Leue, Abdalla Swikir, Fares J. Abu-Dakka