I currently work as a Staff Research Scientist at Boston Dynamics on the Atlas dexterous manipulation team.

Prior to joining Boston Dynamics I worked for 7 years at Agility Robotics where I was one of the first 20 employees and worked on a variety of topics spanning manifold optimization, whole body inverse kinematics, reinforcement learning, contact simulation, subsystem modeling and perception simulation.

Before joining Agility I was an undergrad at Oregon State University where I did research on bipedal locomotion coadvised by Jonathan Hurst and Alan Fern. At OSU I spearheaded the transition of the Dynamic Robotics Laboratory from a hardware focused research group to a software and controls focused group with some of the earliest sim-to-real reinforcement learning results for bipedal running.

My interests broadly lie in the intersection of optimization, simulation and end-to-end learning and control of agile and dexterous humanoid robots.

Experience

2026 - Present

Staff Research Scientist Boston Dynamics

Staff Research Scientist at Boston Dynamics on the Atlas dexterous manipulation team.

2019 - 2026

Robotics Software Engineer Agility Robotics

One of the first 20 employees at Agility Robotics. Most recently on the simulation and innovation teams, previously optimization and controls.

2015 - 2019

Research Assistant Dynamic Robotics Lab

Advisor: Jonathan Hurst, Alan Fern

Planning and deep reinforcement learning based control for bipedal locomotion.

Publications

Towards a Motion Foundation Model: General-Purpose Whole Body Control through Zero-Shot Sim-to-Real Reinforcement Learning

Humanoids (Sim-to-Real Workshop) 2025

Towards a Motion Foundation Model: General-Purpose Whole Body Control through Zero-Shot Sim-to-Real Reinforcement Learning

Jonah Siekman, Helei Duan, Pedro Morais, Chris Paxton, Izzy Brand, Yesh Godse, Alan Fern, Pras Velagapudi, Jonathan Hurst

Iterative Reinforcement Learning Based Design of Dynamic Locomotion Skills for Cassie

CoRL 2019

Iterative Reinforcement Learning Based Design of Dynamic Locomotion Skills for Cassie

Zhaoming Xie, Patrick Clary, Jeremy Dao, Pedro Morais, Jonathan Hurst, Michiel van de Panne

Monte-Carlo Planning for Agile Biped Locomotion

ICAPS 2018

Monte-Carlo Planning for Agile Biped Locomotion

Patrick Clary, Pedro Morais, Alan Fern, Jonathan Hurst

Portfolio

Apex

Apex

PythonPyTorch

A modular pytorch library implementing continuous reinforcement learning algorithms used by the Dynamic Robotics Lab.