Eric Cousineau
Research Engineer at Toyota Research Institute
Cambridge, Massachusetts, United States
Summary
🤩
RockstarEric Cousineau is a research scientist specializing in robotics and imitation learning, currently on Boston Dynamics' Atlas team where he focuses on general manipulation. With nine years in robotics research and engineering, he combines control theory, real-time implementation, and software development to bridge algorithms and hardware. He spent several years at Toyota Research Institute contributing to large behavior models and before that worked on high-impact projects at MathWorks and in the AMBER Lab under Aaron Ames. His open-source contributions include significant improvements to widely used robotics projects like Drake and the MIT underactuated course material, adding visualization, Python 3 compatibility, and MultibodyPlant enhancements. Eric’s background shows a rare blend of hands-on humanoid control experience (Valkyrie, DURUS, PROXI) and backend software architecture, enabling reproducible, testable robotics systems. Based in Cambridge, MA, he brings practical deployment experience alongside research-grade rigor.
9 years of coding experience
Github Skills (21)
c-language10
robotics10
python10
testing10
numpy10
dynamics10
sdformat10
binding10
drake10
pybind1110
cprogramming-language10
architecture9
kinematics9
visualization9
operator-overloading9
Programming languages (20)
Github contributions (5)
RobotLocomotion/drake
Mar 2017 - Jan 2023
Model-based design and verification for robotics.
Role in this project:
Backend Developer & Software Architect
Backend Developer & Software ArchitectContributions:5 releases, 2497 reviews, 1421 commits in 5 years 10 months
Contributions summary:Eric made several contributions to the Drake project, focusing on enhancing the MultibodyPlant functionality. Their work involved enabling optional frame naming within the multibody system and improving the handling of geometry and frames. They added new constructors and methods, and made various corrections to ensure the correct behavior for different data types within the plant. The user also addressed issues in the code, such as fixing potential memory leaks and improving the code's reliability for testing and downstream projects.
roboticsdrake
pybind/pybind11
Aug 2017 - Feb 2022
Seamless operability between C++11 and Python
Role in this project:Back-end Developer & QA Engineer / Test Automation Engineer
Back-end Developer & QA Engineer / Test Automation EngineerContributions:103 reviews, 19 commits, 60 PRs in 4 years 7 months
Contributions summary:Eric primarily contributed to enhancing the pybind11 library's functionality, particularly concerning NumPy integration and operator overloading. Their work included implementing and testing concrete size aliases for NumPy dtypes, adding tests for explicit dtype checks and various operator overloading behaviors. The user also addressed issues related to closures within the eval functionality, demonstrating proficiency in extending the library's capabilities and ensuring its reliability through comprehensive testing.
pythonbindings
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.