Devesh K. Jha

Samsung Research America

Mountain View, California

I build robots that understand contact. I am a Lead Research Scientist in the Robot Intelligence Lab at Samsung Research America, where I am working on several fundamental topics for scalable dexterous manipulation. I am interested in creating multi-modal foundation models for robotics which can understand interaction semantics using vision, touch, force and more.

Previously, I was a Senior Principal Research Scientist at Mitsubishi Electric Research Labs (MERL) in Cambridge, MA, where I led the robotic manipulation effort - setting fundamental research directions and finding avenues for commercializing matured technologies. I also had a brief stint as AI co-founder of a robotics startup, building its manipulation-focused AI strategy.

From Contact to Cognition: I believe robots exist in a physical world where they have to do more than seeing and reacting. I believe true autonomy would come by reasoning about making and breaking contacts with their environment. My research develops physical intelligence for contact-rich robotic manipulation: dexterous manipulation, tactile sensing, contact-implicit trajectory optimization, learning from demonstration, and foundation models for robotics. I have published 80+ papers at venues like ICRA, IROS, RSS, CoRL, T-RO, NeurIPS and ICML, and hold 20 granted or allowed patents.

I received my Ph.D. in Mechanical Engineering from Penn State, along with M.A. degrees in Mathematics and Mechanical Engineering, and my B.E. from Jadavpur University. I serve as Area Chair for NeurIPS, AAAI and ICML, and Associate Editor for IEEE T-ASE, ICRA, and CASE. I have mentored 25+ Ph.D. students and postdocs through research internships, most resulting in publications at top robotics and ML venues.

news

Apr 15, 2026	Joined Samsung Research America as a Lead Research Scientist, working on multi-modal foundation models for robotics and learning from ego-centric videos.
Feb 01, 2026	PPGuide: Steering Diffusion Policies with Performance Predictive Guidance accepted at ICRA 2026.
Nov 15, 2025	Appointed Associate Editor of IEEE Transactions on Automation Science and Engineering (T-ASE).
Nov 01, 2025	Invited talk, “Embracing Contacts for Physical Intelligence,” at Iowa State University.
Oct 15, 2025	Received the MERL Director’s Award for physics-aware ML for anomaly detection in robots.

selected publications

ICRA

PPGuide: Steering Diffusion Policies with Performance Predictive Guidance

Zixing Wang, Devesh K. Jha, and Diego Romeres

In IEEE International Conference on Robotics and Automation (ICRA), 2026

Bib

@inproceedings{wang2026ppguide,
  title = {PPGuide: Steering Diffusion Policies with Performance Predictive Guidance},
  author = {Wang, Zixing and Jha, Devesh K. and Romeres, Diego},
  booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
  year = {2026}
}

RA-L

Simultaneous Extrinsic Contact and In-Hand Pose Estimation via Distributed Tactile Sensing

Mark Merwe, Kei Ota, Nima Fazeli, and 2 more authors

IEEE Robotics and Automation Letters, 2026

Bib

@article{vandermerwe2026tactile,
  title = {Simultaneous Extrinsic Contact and In-Hand Pose Estimation via Distributed Tactile Sensing},
  author = {Van der Merwe, Mark and Ota, Kei and Fazeli, Nima and Berenson, Dmitry and Jha, Devesh K.},
  journal = {IEEE Robotics and Automation Letters},
  year = {2026}
}

CoRL

In-Context Iterative Policy Improvement for Dynamic Manipulation

Mark Merwe and Devesh K. Jha

In Conference on Robot Learning (CoRL), 2025

Bib

@inproceedings{vandermerwe2025incontext,
  title = {In-Context Iterative Policy Improvement for Dynamic Manipulation},
  author = {Van der Merwe, Mark and Jha, Devesh K.},
  booktitle = {Conference on Robot Learning (CoRL)},
  year = {2025}
}

ICRA

Hierarchical Contact-Rich Trajectory Optimization for Multi-Modal Manipulation Using Tight Convex Relaxations

Yuki Shirai, Arvind Raghunathan, and Devesh K. Jha

In IEEE International Conference on Robotics and Automation (ICRA), 2025

Bib

@inproceedings{shirai2025hierarchical,
  title = {Hierarchical Contact-Rich Trajectory Optimization for Multi-Modal Manipulation Using Tight Convex Relaxations},
  author = {Shirai, Yuki and Raghunathan, Arvind and Jha, Devesh K.},
  booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
  year = {2025}
}

IROS

Autonomous Robotic Assembly: From Part Singulation to Precise Assembly

Devesh K. Jha, Kei Ota, Siddarth Jain, and 5 more authors

In IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024

Bib

@inproceedings{jha2024assembly,
  title = {Autonomous Robotic Assembly: From Part Singulation to Precise Assembly},
  author = {Jha, Devesh K. and Ota, Kei and Jain, Siddarth and Yerazunis, Bill and Corcodel, Radu and Shukla, Yash and Bronars, Antonia and Romeres, Diego},
  booktitle = {IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  year = {2024}
}

RSS

Tactile-Filter: Interactive Tactile Perception for Part Mating

Kei Ota, Devesh K. Jha, Hsiao-Yu Tung, and 1 more author

In Robotics: Science and Systems (RSS), 2023

Bib

@inproceedings{ota2023tactilefilter,
  title = {Tactile-Filter: Interactive Tactile Perception for Part Mating},
  author = {Ota, Kei and Jha, Devesh K. and Tung, Hsiao-Yu and Tenenbaum, Joshua B.},
  booktitle = {Robotics: Science and Systems (RSS)},
  year = {2023}
}

ICRA

Tactile RL for Insertion: Generalization to Objects of Unknown Geometry

Siyuan Dong, Devesh K. Jha, Diego Romeres, and 2 more authors

In IEEE International Conference on Robotics and Automation (ICRA), 2021

Bib

@inproceedings{dong2021tactilerl,
  title = {Tactile RL for Insertion: Generalization to Objects of Unknown Geometry},
  author = {Dong, Siyuan and Jha, Devesh K. and Romeres, Diego and Nikovski, Daniel and Rodriguez, Alberto},
  booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
  year = {2021}
}

JDSMC

Imitation of Demonstrations Using Bayesian Filtering with Nonparametric Data-Driven Models

Nurali Virani, Devesh K. Jha, Z. Yuan, and 2 more authors

ASME Journal of Dynamic Systems, Measurement, and Control, 2019

Rudolf Kalman Best Paper Award 2019, ASME DSCD

Bib

@article{virani2019imitation,
  title = {Imitation of Demonstrations Using Bayesian Filtering with Nonparametric Data-Driven Models},
  author = {Virani, Nurali and Jha, Devesh K. and Yuan, Z. and Shekhawat, I. and Ray, Asok},
  journal = {ASME Journal of Dynamic Systems, Measurement, and Control},
  year = {2019},
  note = {Rudolf Kalman Best Paper Award 2019, ASME DSCD}
}