Chaithanya Krishna

Chaithanya Krishna Bodduluri

Senior Robotics & AI Engineer with 7+ years of experience designing and deploying autonomous robot systems from concept to deployment. I'm at Meta, currently with the Agios team (since 2026). From Dec 2023–2025 I was with FAIR in Menlo Park, working at the intersection of robot learning, foundation models (LLM / VLM / VLA), tactile sensing, and dexterous manipulation.

Before Meta, I was a robotics & AI intern at Robert Bosch on the NASA Lunar robot program (built the AprilTag generator on this site as part of that work), and a Robotics Software Engineer with the TCS CTO Team, where I led the system architecture for the MBZIRC 2020 challenge.

News

Experience

Research & Projects

AINA — Dexterity from Smart Lenses

AINA — Dexterity from Smart Lenses

Multi-fingered robot manipulation policies learned from in-the-wild human demonstrations captured with smart-glasses-style egocentric video. Accepted at ICRA 2026 (paper).

DexterityGen

DexterityGen

A foundation controller that produces unprecedentedly dexterous robot hand behaviors — threading syringes, twirling pens, using screwdrivers — from a single learned policy. Accepted at RSS 2025 (paper).

GeoRT — geometric hand retargeting demo

GeoRT — Geometric Retargeting

A principled, ultrafast neural algorithm for retargeting human hand mocap to dexterous robot hands. Trains in 1–2 minutes and deploys in a few lines of Python. Published at IROS 2025 (paper).

Tactile Beyond Pixels (Sparsh-X)

Tactile Beyond Pixels (Sparsh-X)

Multisensory touch representations that fuse Digit 360's tactile image, audio, IMU and pressure into a single embedding for contact-rich robot manipulation. Accepted at CoRL 2025 (Oral) (code, paper).

Sparsh-Skin demo

Sparsh-Skin

Self-supervised perception for tactile-skin covered dexterous hands. Provides embeddings for magnetic-skin tactile data on Allegro / Digit Plexus hands; powers downstream contact-localization and dexterous manipulation. Accepted at CoRL 2025 (code).

Sparsh — self-supervised tactile representations

Sparsh

Self-supervised touch representations for vision-based tactile sensing. Pretrained encoders that transfer across DIGIT, GelSight, and other camera-style tactile sensors. Accepted at CoRL 2024 (code, paper).

Digit 360 multimodal tactile sensor

Digit 360

A modular, high-resolution (~8.3 million taxel) tactile fingertip that captures multi-modal touch signals (force, vibration, audio, proximity, thermal) and processes them on-device. Co-developed at Meta FAIR with GelSight Inc. Lambeta, Wu, Sengül, Most, Black, Sawyer, Mercado, Qi, Bodduluri, …, Calandra (2024).

Meta Digit Plexus tactile hand platform

Meta Digit Plexus

An open hardware-software platform that standardises how tactile sensors (Digit, Digit 360, ReSkin) and end-effectors integrate onto a robot hand. Provides a single-cable hand-to-host interface for fingertips, fingers, and palm sensors, with Python and ROS 2 control. Partnered with Wonik Robotics on a next-gen Allegro Hand.

AprilTag

AprilTag Generator

A small browser tool to generate AprilTag 2 and AprilTag 3 markers of any family and size, ready to print or save as SVG. Useful when you need fiducials in a hurry for a robotics demo.

MBZIRC 2020

MBZIRC 2020 — Autonomous Wall Building with a Swarm of UAVs

Designed the system architecture and coding stack for a swarm of drones that autonomously picked, transported, and placed bricks to build a wall. Implemented perception, localization, and computer-vision pipelines for Challenge 2 of the Mohamed Bin Zayed International Robotics Challenge.

Publications

Selected works. See Google Scholar for the full list.

Dexterity from Smart Lenses: Multi-Fingered Robot Manipulation with In-the-Wild Human Demonstrations

I. Guzey, H. Qi, J. Urain, C. Wang, J. Yin, K. Bodduluri, M. Lambeta, L. Pinto, et al.

IEEE International Conference on Robotics and Automation (ICRA), 2026.  [project, arXiv]

Sparsh-Skin: Self-Supervised Perception for Tactile Skin Covered Dexterous Hands

A. Sharma, C. Higuera, C. K. Bodduluri, Z. Liu, T. Fan, T. Hellebrekers, et al.

Conference on Robot Learning (CoRL), 2025.  [code, arXiv]

Tactile Beyond Pixels: Multisensory Touch Representations for Robot Manipulation

C. Higuera, A. Sharma, T. Fan, C. K. Bodduluri, B. Boots, M. Kaess, M. Lambeta, et al.

Conference on Robot Learning (CoRL), 2025.  [project, code, arXiv]

Geometric Retargeting: A Principled, Ultrafast Neural Hand Retargeting Algorithm

Z. H. Yin, C. Wang, L. Pineda, K. Bodduluri, T. Wu, P. Abbeel, M. Mukadam

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2025.  [paper]

DexterityGen: Foundation Controller for Unprecedented Dexterity

Z. H. Yin, C. Wang, L. Pineda, F. Hogan, K. Bodduluri, A. Sharma, P. Lancaster, et al.

Robotics: Science and Systems (RSS), 2025.  [arXiv]

Sparsh: Self-Supervised Touch Representations for Vision-Based Tactile Sensing

C. Higuera, A. Sharma, C. K. Bodduluri, T. Fan, P. Lancaster, M. Kalakrishnan, et al.

Conference on Robot Learning (CoRL), 2024.  [arXiv]

PARA-FM: Graphical User Interface for Emergency Flood Planning and Management

U. Rai, D. Pushp, A. Choudhary, A. Kumar, B. Chaitanya, R. Lima, et al.

5th World Congress on Disaster Management, Vol. II, pp. 100–106, 2022.  [chapter]

A Bifurcated Gain Adaptation Strategy in Sliding Mode Control with Chattering Alleviation for High Fidelity Operations

N. Sridhar, N. S. Abhinay, B. C. Krishna, K. Das

American Control Conference (ACC), pp. 2139–2144, 2021.  [paper]