AI People Wiki

Builders & open source

Emanuel Todorov

Research and development, Roboti LLC

  • Research and development, Roboti LLC13

Emanuel (Emo) Todorov works on research and development at Roboti LLC, the company that developed the MuJoCo physics simulator for multi-body dynamics with contact. Todorov's CV lists faculty roles in Cognitive Science at UC San Diego from 2002 to 2009 and as Associate/Affiliate Professor of Applied Mathematics and Computer Science & Engineering at the University of Washington from 2009; the Movement Control Laboratory's site is now an archive. The lab built MuJoCo for model-based control, and Roboti's page notes that DeepMind acquired MuJoCo in October 2021 and released it as open source under the Apache 2.0 license. Todorov also co-authored a 2017 paper on learning dexterous manipulation with deep reinforcement learning and demonstrations.12345

At a glance

Current role
Research and development
Company
Roboti LLC
Sector
Company
Education
MIT, UCL, UC San Diego, USC, UW
Work listed
2
Honors
2
Sources cited
5
Profile reviewed

Career

Todorov earned a PhD in cognitive neuroscience at MIT in 1998 and was a postdoctoral fellow at the Gatsby Computational Neuroscience Unit at University College London from 1999 to 2001. After a year as a research scientist at the Alfred Mann Institute for Biomedical Engineering at the University of Southern California, Todorov joined UC San Diego's Department of Cognitive Science in 2002.2

Todorov founded the Movement Control Laboratory at UC San Diego in 2002 and moved it to the University of Washington's Department of Computer Science and Engineering in 2009. The archival lab page says that its members have since graduated and that Todorov now works on research and development at Roboti LLC.12

Research

The Movement Control Laboratory studied the control of complex movement in animals and robots, treating both as optimization problems and developing methods for optimal control and learning. Its early work modeled human sensorimotor control, including a 2002 Nature Neuroscience paper on optimal feedback control as a theory of motor coordination.14

Later projects included linearly-solvable optimal control, contact-invariant trajectory optimization for generating complex behaviors automatically, real-time model-predictive control for humanoid robots, and the design of a pneumatically actuated 20-degree-of-freedom robotic hand.4

MuJoCo

The lab built MuJoCo, a physics engine for model-based control that combines recursive algorithms in generalized coordinates with contact solvers based on the lab's own formulations of contact physics; it was described at IROS 2012. Roboti LLC offered MuJoCo as a commercial product from 2015 to 2021. In October 2021 DeepMind acquired MuJoCo and released it as open source under the Apache 2.0 license.43

Work and contributions

  • MuJoCo Open-source code

    The lab's projects page describes MuJoCo as "our physics engine designed for model-based control" and lists the IROS 2012 paper "MuJoCo: A physics engine for model-based control"; Roboti LLC developed it and offered it as a commercial product from 2015 to 2021 before DeepMind made it open source under Apache 2.0.43

  • Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations Published research

    Co-author, with Aravind Rajeswaran, Vikash Kumar, Abhishek Gupta, Giulia Vezzani, John Schulman and Sergey Levine, of the 2017 paper on deep reinforcement learning for multi-fingered hands.5

Timeline

  1. Completed a PhD in cognitive neuroscience at the Massachusetts Institute of Technology.2
  2. Became a postdoctoral fellow at the Gatsby Computational Neuroscience Unit, University College London.2
  3. Joined the Alfred Mann Institute for Biomedical Engineering at the University of Southern California as a research scientist.2
  4. Became an assistant professor of cognitive science at UC San Diego and founded the Movement Control Laboratory.21
  5. Promoted to associate professor in UC San Diego's Department of Cognitive Science.2
  6. Moved the Movement Control Laboratory to the University of Washington, joining Applied Mathematics and Computer Science & Engineering.21
  7. Presented the paper MuJoCo, a physics engine for model-based control, at IROS 2012.4
  8. MuJoCo became available as a commercial product from Roboti LLC.3
  9. Co-authored a paper on learning dexterous manipulation with deep reinforcement learning and demonstrations.5
  10. DeepMind acquired MuJoCo and released it as open source under the Apache 2.0 license.3

Other documented roles

Roles the sources document that are not confirmed as current.

Honors and recognition

  • Sloan Research Fellowship in Neuroscience · Alfred P. Sloan Foundation, 20042
  • Howard Hughes Predoctoral Fellowship · Howard Hughes Medical Institute, 19942

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

5 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. Emo Todorov — Movement Control Laboratory (archival page, formerly hosted at the University of Washington)Emo Todorov · Official page · accessed Oct 21, 2026
  2. Emo Todorov — academic CVEmo Todorov · Official page · accessed Oct 21, 2026
  3. Roboti LLC — MuJoCoRoboti LLC · Official page · accessed Oct 21, 2026
  4. Emo Todorov — lab projects (MuJoCo and model-predictive control)Emo Todorov · Official page · accessed Oct 21, 2026
  5. Learning Complex Dexterous Manipulation with Deep Reinforcement Learning and Demonstrations (arXiv 1709.10087)arXiv · Paper · published Sep 28, 2017 · accessed Oct 21, 2026