Aditya Narendra

Robotics Researcher

Hi, I am Aditya.

I make Robots smarter and safer.

  • RoleMaster's student, MBZUAI Robotics Department
  • FocusSmarter, safer and more precise manipulation control
  • BasedAbu Dhabi, UAE
  • Emailaditya.narendra@mbzuai.ac.ae
Download CV Portfolio

aditya $ cat about.txt

I am a robotics and reinforcement learning researcher working on manipulation problems, in particular making the control smarter, safer and more precise.

I started at MBZUAI as a Research Engineer and am now a master's student in the Robotics Department and AIRoCS Lab. Previously, I worked with the Center for Cognitive Modeling at MIPT, where I completed my BSc in Applied Mathematics and Computer Science.

aditya $ ls research_interests/

  • Multi-task reinforcement learning
  • Model-based policy optimization
  • Knowledge graphs for manipulation
  • Symbolic planning and certified control
  • Simulation environments and benchmarking
  • Franka, UR5, UGV and mobile manipulation systems

Publications

Selected publications

Manuscript 2026

M4PO

Massively Multi-Task Multi-Embodiment Model-Based Policy Optimization across robot embodiments.

Project Page Paper Code

IROS 2025

M3PO

Massively Multi-Task Model-Based Policy Optimization for scalable robotic control.

Project Page arXiv Code

ICRA 2026

KG-M3PO

Knowledge-guided manipulation using online scene graphs and multi-task reinforcement learning.

Project Page arXiv Code

More publications

Timeline

Recent research path

2026

KG-M3PO was accepted to ICRA 2026; started at MBZUAI as a Research Engineer and then became a master's student in August.

2025

M3PO published at IROS 2025; graduated from MIPT after a thesis on multi-task RL for manipulation.

2024

Research Engineer at the Center for Cognitive Modeling, MIPT.

2023

Research intern at MIPT, working across control, trajectory optimization, and robot learning.