Gautam Salhotra

Gautam Salhotra

PhD candidate in robotics

University of Southern California

I am a PhD candidate in computer science (robotics) at the University of Southern California (USC), advised by Prof. Gaurav Sukhatme. Broadly, I work on combining planning algorithms with learning to achieve greater autonomy for robots, with experience in object manipulation, control theory, and deep reinforcement learning.

Download my résumé (2 page) and CV (full).

  • Robot manipulation
  • Deformable Objects
  • Learning & Optimization
  • MSc in Computer Science (Robotics & Perception), 2018

    Georgia Tech

  • MSc in Mechanical Engineering (Control Theory), 2012

    University of Texas at Austin

  • Mechanical Engineering, 2010

    Indian Institute of Technology Bombay (IIT Bombay)


  • Sep 2023 The USC Robotics Seminar (URoS) is back for regular seminars from USC researchers this Fall.

  • Aug 2023 Our paper on cross-embodied learning for deformable object manipulation has been accepted to CoRL 2023! See you in Atlanta!

  • Jun 2023 Our paper on Learned Parameter Selection for Robotic Information Gathering has been accepted to IROS 2023! See you in Detroit!

  • May 2023 I started as a PhD resident at Intrinsic LLC (Alphabet).

  • Mar 2023: I completed my PhD thesis proposal.

  • Feb 2023: I am organizing the USC Robotics Seminar (URoS) along with Prof. Somil Bansal, a monthly seminar for all robotics researchers at USC to present and discuss their work.

  • Oct 2022: I presented our work on learning deformable object manipulation from expert demonstrations at IROS 2022 and the workshop on RObotic MAnipulation of Deformable Objects (ROMADO-SI).

  • Sep 2022: Two papers will be presented at the Southern California Robotics Symposium 2022.

    • Learning deformable object manipulation from demonstrations (DMfD)
    • Guided Learning of Robust Hurdling Policies with Curricular Trajectory Optimization (CTO-RL)
  • Jun - Sep 2022: I will be an Applied Scientist intern at Amazon Robotics.

  • Jun 2022: Our paper on learning deformable object manipulation from expert demonstrations has been accepted to both IEEE RA-L and IEEE IROS 2022. See you in Kyoto! (Website || RA-L link)

Work Experience

Applied Scientist Intern
Jun 2022 – Sep 2022 Greater Boston area
Researched & developed manipulation policies for delicate items.
Robotics Research Intern
Bosch Research
May 2019 – Aug 2019 California, Bay area
  • Reinforcement Learning for peg insertion tasks (environments, learning and classical control methods)
  • Developed ROS package to deploy a learned algorithm, tested on robot hardware.
Senior Software Controls Engineer
Jan 2016 – Jun 2018 Greater Boston area
  • Implemented object manipulation algorithms to pick & place cases in warehouse storage and retrieval systems (C++, python).
  • Worked on low-level controllers for actuator performance and stall detection.

Selected Publications

See Google Scholar for a full list

(2022). Learning Deformable Object Manipulation from Expert Demonstrations. In IEEE RA-L, IROS ‘22.

PDF Cite Website Code

(2022). Guided Learning of Robust Hurdling Policies with Curricular Trajectory Optimization. In SoCal Robotics Symposium ‘22.

PDF Cite Website

(2021). Adaptive Sampling using POMDPs with Domain-Specific Considerations. In ICRA ‘21.


(2020). Motion Planner Augmented Reinforcement Learning for Robot Manipulation in Obstructed Environments. In CoRL ‘20.

Cite Website PDF Code




  • Chaired ICRA 2021 session on ‘Field Robotics: Control’