Physics simulation / robotics / digital humans

Jeongseok "JS" Lee

I build physics simulation systems for robotics, digital humans, and AR/VR, with a focus on multibody dynamics, contact, humanoid robot simulation, and practical open-source software.

Software CV

Portrait of Jeongseok Lee — **Robotics Software Engineer**Reality Labs Research | Meta Robotics Studio

Current direction

Simulation systems that bridge research and real-time platforms.

Software engineer and robotics scientist working at the boundary between simulation research and production systems. Current work spans physics-based simulation for avatars and full-stack simulation frameworks for humanoid robots. Long-term open-source work includes leading DART, the Dynamic Animation and Robotics Toolkit.

The throughline is practical physics software: robust enough for research, maintainable enough for production teams, and clear enough to serve open communities.

Software

Robotics software, simulation infrastructure, and digital-human systems.

Public engineering work organized around durable robotics libraries, research-facing simulation systems, and tools that make implementation judgment visible.

Lead developer, Nov 2012 - present

DART

Dynamic Animation and Robotics Toolkit

Research-focused C++ physics engine for robotics, animation, and machine learning, with Python bindings and long-running use in robotics simulation research.

Kinematics, dynamics, collision detection, optimization, and articulated rigid-body simulation.
JOSS publication and Grand Prize, OSS World Challenge 2016.
Integrated into Gazebo workflows for DARPA Robotics Challenge research.

Website GitHub Paper

Research software, Meta

MHR

Momentum Human Rig

An anatomically inspired parametric full-body digital human model for computer graphics and computer vision research.

Includes a skeletal model, skinned mesh levels of detail, body blendshapes, pose correctives, and facial blendshape model.
Connects Jeongseok's simulation background to current work on digital humans and avatars.

GitHub PDF

Research software, Meta

Momentum

Human kinematics and optimization

A library for human kinematic motion and numerical optimization solvers for applying and analyzing human motion.

Public software around kinematic motion, numerical optimization, and practical tooling.
Complements MHR as part of a broader digital-human research stack.

GitHub

Contributor, 2016 - 2018

AIKIDO

Motion planning and decision making

Robotic manipulation framework used by the Personal Robotics Lab, with work spanning inverse kinematics, collision detection, and robot dynamics.

Deployed to lab robot platforms including HERB and ADA.
Connected planning, dynamics, and system integration in manipulation workflows.

GitHub

Abstract FCL collision detection graphic

Contributor

FCL

Flexible Collision Library

Contributions to collision detection infrastructure used across robotics, motion planning, and simulation workflows.

Worked on collision detection infrastructure for practical robotics software stacks.
Connected geometry, contact, and simulation needs across planning and dynamics workflows.

FCL

Contributor

Gazebo

Robot simulation ecosystem

Integration work in the GUI-based robot simulation ecosystem used by robotics researchers and developers.

Helped bring DART into established Gazebo simulation workflows.
Connected physics engine work to robot simulation tools used by research teams.

Gazebo

Meta public work

Work-specific open-source maintenance and packaging.

Meta-related public repositories, package recipes, and ecosystem maintenance are grouped separately from personal open-source projects.

Meta public ecosystem

jeongseok-meta

Meta-specific public GitHub account

Public forks, package recipes, and Meta-related open-source maintenance around digital humans, simulation, visualization, and research tooling.

Keeps work-specific public maintenance separate from the personal jslee02 account.
Includes public work around MHR, Momentum, DART, viser, and supporting C++/Python infrastructure.

GitHub

Packaging / ecosystem

conda-forge feedstock maintenance

Packaging infrastructure for research software

Tracked conda-forge inventory for robotics, vision, geometry, visualization, ML, and simulation-adjacent packages.

58 feedstocks created, 11 co-maintained, 1 additional feedstock contributed to, and 70 total feedstocks listed.
Includes packages such as MHR, Momentum, DART, viser, nerfstudio, gpytoolbox, mmpose, and AdaptiveCpp.

Inventory Dashboard

Curated lists

Awesome lists maintained separately from project work.

These collections are useful discovery surfaces for robotics, collision detection, graphics, GPGPU, ECS, and C++/Python tooling.

Maintainer

Awesome Robotics Libraries

Robotics software map

Curated robotics libraries and software across planning, dynamics, perception, control, simulation, and tooling.

GitHub

Maintainer

Awesome Collision Detection

Geometry and proximity queries

Curated collision detection libraries and resources for robotics, simulation, motion planning, and graphics.

GitHub

Maintainer

Awesome Graphics Libraries

3D graphics resources

Curated 3D graphics libraries and resources relevant to simulation, visualization, rendering, and tooling.

GitHub

Maintainer

Awesome Entity Component System

ECS libraries and resources

Curated Entity-Component-System libraries and resources for scalable simulation, game, and tool architectures.

GitHub

Maintainer

Awesome GPGPU

CUDA, OpenCL, Vulkan, and compute

Curated GPGPU resources for heterogeneous compute work spanning CUDA, OpenCL, Vulkan, and related tooling.

GitHub

Maintainer

Awesome Robotics Simulation

Multibody dynamics papers

Curated resources for multibody dynamics simulation papers and robotics simulation research context.

GitHub

Maintainer

Awesome C++ Python Binding Generators

Binding generator survey

Curated automatic Python binding generators for C++ projects, useful for simulation and robotics libraries with Python APIs.

GitHub

Research

Physics simulation, motion planning, and robotics systems.

Publications and research outputs spanning multibody dynamics, contact, differentiable physics, robotic manipulation, and digital humans.

Publications

Selected papers

MHR: Momentum Human Rig

A. Ferguson et al. (incl. Jeongseok Lee)

arXiv, 2025

PDF

Fast and Feature-Complete Differentiable Physics for Articulated Rigid Bodies with Contact

Keenon Werling, Dalton Omens, Jeongseok Lee, Ioannis Exarchos, C. Karen Liu

Robotics: Science and Systems, 2021

PDF

Bayesian Policy Optimization for Model Uncertainty

Gilwoo Lee, Brian Hou, Aditya Mandalika, Jeongseok Lee, Siddhartha S. Srinivasa

International Conference on Learning Representations, 2019

PDF

DART: Dynamic Animation and Robotics Toolkit

Jeongseok Lee, Michael X. Grey, Sehoon Ha, Tobias Kunz, Sumit Jain, Yuting Ye, Siddhartha S. Srinivasa, Mike Stilman, C. Karen Liu

Journal of Open Source Software, 2018

DOI PDF

Sampling of Pareto-Optimal Trajectories using Progressive Objective Evaluation in Multi-Objective Motion Planning

Jeongseok Lee, Daqing Yi, Siddhartha S. Srinivasa

IEEE/RSJ International Conference on Intelligent Robots and Systems, 2018

PDF

A Linear-Time Variational Integrator for Multibody Systems

Jeongseok Lee, C. Karen Liu, Frank C. Park, Siddhartha S. Srinivasa

Algorithmic Foundations of Robotics, 2016

PDF

Orienting in Mid-air through Configuration Changes to Achieve a Rolling Landing for Reducing Impact after a Fall

Jeffrey T. Bingham, Jeongseok Lee, Ravi N. Haksar, Jun Ueda, C. Karen Liu

IEEE/RSJ International Conference on Intelligent Robots and Systems, 2014

A Discrete Mechanics-Based Recursive Algorithm for Rigid Multibody Systems

Jeongseok Lee, Jisoo Hong, Frank C. Park

Japan-Korea Joint Symposium on Dynamics and Control, 2013

Optimal Landing Motions for Legged Robots with Variable Stiffness Joints

Jeongseok Lee, Frank C. Park

International Conference on Advanced Mechatronics, 2010

PDF

srLib: The Next Standard in Robotics Physics Engines

Jeongseok Lee, Frank C. Park

International Conference on Ubiquitous Robots and Ambient Intelligence, 2009

Patents

Robotic workcell for interacting with goods to person systems

A. Stubbs, A. J. Parness, S. B. Karumanchi, J. R. Hooks, P. Owan, R. B. Warrier, A. C. C. Lopez, J. Lee, C. Salisbury

US Patent 12,013,686

PDF

Robotic picking assemblies configured to grasp multiple items

J. Lee, A. J. Parness, A. Stubbs

US Patent 11,478,942

PDF

Awards and service

Awards

Grand Prize, OSS World Challenge 2016, The Ministry of Science, ICT and Future Planning of Korea, with DART
Outstanding video paper award, URAI 2009, for srLib

Talks

Amazon Machine Learning Conference, invited talk, Apr 2018

Peer review

Journal of Mechanisms and Robotics, reviewer, 2012
IEEE Robotics and Automation Letters, reviewer, 2018
IEEE International Conference on Robotics and Automation, reviewer, 2015, 2017-2018
IEEE/RSJ International Conference on Intelligent Robots and Systems, reviewer, 2017-2018
Journal of Open Source Software, reviewer, 2017

Jeongseok Lee

Robotics software engineer focused on physics simulation for robotics, humanoid robots, digital humans, and AR/VR systems.

Download PDF CV

Technical experience

Career

Apr 2022 - Present

Staff Research Engineer | Staff Software Engineer

Reality Labs Research | Meta Robotics Studio, Meta Platforms · Redmond, WA

Developed physics-based simulation for avatars in AR/VR applications to enhance immersion and realism in virtual experiences.
Building a full-stack simulation framework for humanoid robots, encompassing physics engines, rendering, sensor simulation, and sim-to-real transfer.

Feb 2019 - Apr 2022

Robotics Software Engineer | Applied Scientist

Robotics AI, Amazon Robotics · Seattle, WA

Developed scalable simulation infrastructure for robotic workcells.
Developed physics-based simulation for articulated robot systems.

Jun 2018 - Sep 2018

Software Engineering Intern

X, formerly Google X · Mountain View, CA

Developed efficient robot simulation software for parallel mechanism systems in contact-rich environments.

Jun 2017 - Aug 2017

Visiting Scientist

Personal Robotics Lab, University of Washington · Seattle, WA

Worked on robotic manipulation, complex dynamical systems, motion planning, and system integration.

Jan 2016 - Jun 2017

Senior Software Engineer

Personal Robotics Lab, Carnegie Mellon University · Pittsburgh, PA

Developed and managed AIKIDO, a motion planning and decision-making library for robotic manipulation.
Worked on inverse kinematics, collision detection, robot dynamics, and energy-conservative integration methods.

Mar 2013 - Jan 2016

Research Scientist

Graphics Lab, Georgia Institute of Technology · Atlanta, GA

Developed DART with kinematics, dynamics, collision detection, and optimization.
Integrated DART into Gazebo for DARPA Robotics Challenge workflows.

Education

M.S. in Computer Science and Engineering

University of Washington

Sep 2017 - Dec 2018

Advised by Siddhartha S. Srinivasa

Ph.D. candidate (ABD) in Mechanical and Aerospace Engineering

Seoul National University

Mar 2007 - Feb 2013

Advised by Frank C. Park

B.S. in Naval Architecture and Ocean Engineering

Seoul National University

Mar 2002 - Aug 2006

Advised by Kyu-Yeul Lee

Skills

Languages

Simulation

Systems

Contact

Reach out about robotics, simulation, digital humans, or open-source software.

The fastest public links are GitHub, LinkedIn, Google Scholar, and email.

GitHubhttps://github.com/jslee02 Meta GitHubhttps://github.com/jeongseok-meta LinkedInhttps://www.linkedin.com/in/jeongseok-lee/Google Scholarhttps://scholar.google.com/citations?user=DJ6mzSUAAAAJ&hl=en ORCIDhttps://orcid.org/0000-0002-2146-7502 Xhttps://x.com/jeongseok_lee_Emailjslee02@gmail.com