Narayanan PP

Graduate Student at RI, CMU

prof_pic.jpg

I am an incoming MS in Robotic Systems Development (MRSD) student at the Robotics Institute, Carnegie Mellon University. My research interests lie at the intersection of robot learning, embodied AI, and language-informed robotics, with a focus on building generalizable systems that can rapidly and autonomously acquire diverse skills in the real world.

Previously, I was a researcher at the Indian Institute of Science (IISc), Bangalore, where I was affiliated with the Stochastic Robotics Lab, led by Dr. Shishir Kolathaya.

My recent project at IISc, GroQLoCO (Generalist and Robot-agnostic Quadruped Locomotion Control using Offline Datasets), involved developing generalist locomotion frameworks that enable zero-shot transfer across diverse quadruped morphologies and terrains using offline multi-behavior learning. This line of work focuses on Imitation Learning for scalable and morphology-agnostic control.

I have also worked on sim-to-real transfer for quadruped locomotion, designing reinforcement learning-based controllers that generalize across environments and terrains, and deploying them on custom hardware platforms as well as the Unitree Go1 robot.

While at IISc, I collaborated with the Improbable AI Lab, led by Dr. Pulkit Agrawal at MIT, on deep RL-based locomotion research for custom quadrupeds.

I have collaborated with Mr. Anantharaman PN, Founder & CEO of JN Research Labs, Bangalore, on Generative AI for structured data analytics. This included research on tabular Q&A, summarization, and building semantic representations for structured data. During my earlier internship at JN Research Labs, I developed a joystick-based teleoperation interface for wheeled surveillance robots.

I graduated with a B.Tech in Mechanical Engineering from National Institute of Technology, Bhopal, India (NIT Bhopal) in 2024.

Selected Publications

  1. groqloco.gif
    GRoQ-LoCO: Generalist and Robot-agnostic Quadruped Locomotion Control using Offline Datasets
    Narayanan PP, Sarvesh Prasanth Venkatesan, Srinivas Kantha Reddy, and 1 more author
    2025
    Under review at CoRL 2025
  2. hysem.png
    HySem: A context length optimized LLM pipeline for unstructured tabular extraction
    Narayanan PP, and Anantharaman Palacode Narayana Iyer
    2024
    TRL workshop at NeurIPS 2024