Aditya Penumarti

Ph.D. Candidate @ University of Florida

Adi.jpeg

Me between a felled giant sequoia at Calveras Big Trees State Park, CA

I am a Ph.D. candidate at the University of Florida in the Mechanical and Aerospace Engineering Department, where I study how robots can make good decisions when they can’t be sure where they are or what’s around them. My research focuses on planning under uncertainty in GPS-denied environments, such as underwater or indoor environments. The overarching theme is active perception: my algorithms treat sensing as active, allowing them to choose to gather the information a robot needs to stay localized. This includes belief-space planning, uncertainty-aware representations of the environment, and information-constrained planning that decides how much sensing is enough. The algorithms I design have been tested on underwater and ground robots with support from the Air Force Research Laboratory (AFRL).

I am advised by Prof. Jane Shin. I received my M.Sc from the University of Florida, working with Prof. Carl Crane and Prof. Michael Griffis. I received my B.Sc. from North Carolina State University.

I am deeply interested in teaching and mentoring. I co-developed an autonomous ground vehicle class at the University of Florida, built on 1/10 vehicle platforms, and I have mentored five undergraduate and master’s students.

News

Jul 31, 2026 Our paper “Calibration and Comparative Analysis of Forward-Looking Sonar and 3D Sonar for Enhanced Underwater Object Recognition” was accepted to the IEEE OCEANS 2026 conference!
Mar 26, 2026 I passed my proposal exam and am a Ph.D. Candidate.
Jan 31, 2026 Our paper “Real-time Uncertainty-Aware Motion Planning for Magnetic-based Navigation” was accepted for publication in IEEE Access!

Selected Publications

  1. ACCESS
    InfoMagNav.gif
    Real-time Uncertainty-Aware Motion Planning for Magnetic-based Navigation
    Aditya Penumarti, Humberto Ramos, Kristy Waters, and 2 more authors
    IEEE Access, Jan 2026
  2. CoRL
    Belief-Conditioned One-Step Diffusion: Real-Time Trajectory Planning with Just-Enough Sensing
    Gokul Puthumanaillam, Aditya Penumarti, Manav Vora, and 5 more authors
    In Conference on Robot Learning, Sep 2025