Evan Drumwright

Evan Drumwright


I graduated in 2007 from The University of Southern California with a Ph. D. in Computer Science (concentrating on robotics) under Maja Mataric'. I graduated from the University of Memphis (which has one of the top mathematics programs in the US) with a B.S. in Mathematics and a B.S. in Computer Science (1999).

I discovered the first polynomial time model for simultaneous impacts between rigid bodies, the first quadratic time algorithm for inverting generalized inertia matrices, and- with my student Sam Zapolsky)- the first algorithm for computing inverse dynamics under the rigid contact model.

I am currently CEO of a startup, Dextrous Robotics. We're in "stealth mode" and not sharing any information publicly just yet.

Last position: I was a senior research scientist at Toyota Research Institute, where I developed the Drake simulator. I also developed a really neat contact model with Ryan Elandt, Michael Sherman (Sherm), and Andy Ruina.

Second to last position: I was faculty at GWU, where my work was funded by NSF, Willow Garage, Open Source Robotics Foundation/NASA, and Hewlett Packard.


I ran the Positronics Lab (Planning Optimization and Simulation for Robotics) at GWU. Please see my lab web page for publications, students, facilities, and joining my research group.


I am interested in the computational aspects of simulating the natural world. Since 2005, I have focused on rigid body and multi-rigid body simulation, inverse dynamics, and modeling contact. I typically try to tie my work in these areas to my interests in robotics.

I have been fortunate enough to grow up in the Information Age. I hope to live long enough to see the first wave of the Robotics Age, which will automate the physical world like computers have automated the informational one. I am not very interested in robots doing my housework. I am very interested in robots that transform society, like those that help elderly people retain physical independence, serve as trustworthy playmates and guardians for our children, and watch over our homes when we're away.

When I was a teenager, I guessed that the most difficult part of creating such robots would be making them conscious. Pop culture seems to believe this too. I now think the biggest challenges lie in the areas of locomotion and physical interaction. My interests in robotics and simulating the dynamics of the natural world have led to my recent research in legged robot locomotion with Ph. D. student Sam Zapolsky and human interaction with Ph. D. students Josh Lurz and Roxana Leontie. I hope to one day make a robot with the athletic abilities of squirrels and goats.

Contact info

The most reliable way to contact me is via email: edrumwri at gmail dot com.