Research field
Learning from demonstration and physical interaction
Robots acquire skills from human demonstration or from their own exploration. These skills then transfer across objects, goals and embodiments.
About this field
A robot has to come to have a skill in the first place. It can be shown by a person, or discovered through its own exploration and contact with the world. The skill must then survive the move to new objects, new goals and different bodies.
People

Nilay Kushawaha
The research focuses on continual learning for robot manipulation. It also studies cross-embodiment transfer, so that skills carry over between different robots.
Nilay Kushawaha is a Postdoctoral Associate at NYU Abu Dhabi, with a Ph.D. in Biorobotics and AI from the Scuola Superiore Sant'Anna, Italy.
His research focuses on continual learning, brain-inspired AI, and adaptive robot control. He develops algorithms that let robots keep learning and adapting across their lifetime, the way humans do. His work spans both soft and rigid robotic systems, combining machine learning, robotics, and physics to build more intelligent and autonomous robots.
Keywords: Continual Learning, Incremental Learning, Learning from Demonstrations, Reinforcement Learning, Soft & Rigid Robotics, Sensor Fusion.

Omar Mostafa
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.
Publications since 2024
Safe and Stable Neural Network Dynamical Systems for Robot Motion Planning
IEEE Robotics and Automation Letters (RA-L), vol. 11, no. 3, pp. 3111–3118
@article{binny2026safe, author = {Binny, A. E. and Anand, M. and Kussaba, H. T. M. and Chen, L. and Agrawal, S. and Abu-Dakka, F. J. and Swikir, A.}, title = {{Safe and Stable Neural Network Dynamical Systems for Robot Motion Planning}}, journal = {IEEE Robotics and Automation Letters}, year = {2026}, volume = {11}, number = {3}, pages = {3111--3118}, doi = {10.1109/LRA.2026.3655207}, eprint = {2511.20593}, archivePrefix = {arXiv} }MV-UMI: A Scalable Multi-View Interface for Cross-Embodiment Learning
IEEE Access
@article{rayyan2026mvumi, author = {Rayyan, O. and Abanes, J. and Hafez, M. and Tzes, A. and Abu-Dakka, F.}, title = {{MV-UMI: A Scalable Multi-View Interface for Cross-Embodiment Learning}}, journal = {IEEE Access}, year = {2026}, volume = {14}, pages = {104545--104555}, doi = {10.1109/ACCESS.2026.3709982}, eprint = {2509.18757}, archivePrefix = {arXiv} }MeshDMP: Motion Planning on Discrete Manifolds using Dynamic Movement Primitives
IEEE International Conference on Robotics and Automation (ICRA)
@inproceedings{dallevedove2025meshdmp, author = {Dalle Vedove, M. and Abu-Dakka, F. J. and Palopoli, L. and Fontanelli, D. and Saveriano, M.}, title = {{MeshDMP: Motion Planning on Discrete Manifolds using Dynamic Movement Primitives}}, booktitle = {IEEE International Conference on Robotics and Automation (ICRA)}, year = {2025}, pages = {895--901}, publisher = {IEEE}, address = {Atlanta, GA, USA}, doi = {10.1109/ICRA55743.2025.11128556}, eprint = {2410.15123v2}, archivePrefix = {arXiv} }Learning Periodic Skills for Robotic Manipulation: Insights on Orientation and Impedance
Robotics and Autonomous Systems (RAS)
@article{abudakka2024learning, author = {Abu-Dakka, F. and Saveriano, M. and Peternel, L.}, title = {{Learning Periodic Skills for Robotic Manipulation: Insights on Orientation and Impedance}}, journal = {Robotics and Autonomous Systems}, year = {2024}, volume = {180}, pages = {104763}, doi = {10.1016/j.robot.2024.104763} }A Unified Formulation of Geometry-Aware Discrete Dynamic Movement Primitives
Neurocomputing
@article{abudakka2024unified, author = {Abu-Dakka, F. J. and Saveriano, M. and Kyrki, V.}, title = {{A Unified Formulation of Geometry-Aware Discrete Dynamic Movement Primitives}}, journal = {Neurocomputing}, year = {2024}, volume = {598}, pages = {128056}, doi = {10.1016/j.neucom.2024.128056}, eprint = {2203.03374v3}, archivePrefix = {arXiv} }Fusion Dynamical Systems with Machine Learning in Imitation Learning: A Comprehensive Overview
Information Fusion
@article{hu2024fusion, author = {Hu, Y. and Abu-Dakka, F. J. and Chen, F. and Luo, X. and Li, Z. and Knoll, A. and Ding, W.}, title = {{Fusion Dynamical Systems with Machine Learning in Imitation Learning: A Comprehensive Overview}}, journal = {Information Fusion}, year = {2024}, volume = {108}, pages = {102379}, doi = {10.1016/j.inffus.2024.102379}, eprint = {2403.19916}, archivePrefix = {arXiv} }G3VLA: Geometric inductive bias for Vision-Language-Action Models
Preprint arXiv preprint — not yet peer-reviewed
@misc{peng2026g3vla, author = {Peng, Yue and Zhao, Yongzhe and Habuda, Artur and Pham, Khuyen and Zhu, Yanheng and Le, Tran Nguyen and Abu-Dakka, Fares and Guo, Li}, title = {{G\textsuperscript{3}VLA: Geometric inductive bias for Vision-Language-Action Models}}, year = {2026}, eprint = {2606.24472}, archivePrefix = {arXiv}, note = {Preprint, not yet peer-reviewed}, url = {https://arxiv.org/abs/2606.24472} }

