David Stavens
Papers
11
Total Citations
4,078
H-Index
10
About
David Stavens is a pioneering roboticist whose work on autonomous vehicles helped define the modern self-driving car. As a key member of the Stanford Racing Team, he made foundational contributions to the DARPA Grand and Urban Challenges. Stavens was instrumental in developing **Stanley**, the robot that won the 2005 DARPA Grand Challenge (2,109 citations), and **Junior**, Stanford’s entry in the 2007 Urban Challenge (1,004 citations). These vehicles demonstrated that AI and machine learning could enable high-speed desert driving and complex urban navigation, including lane changes and interaction with other traffic. Beyond these landmark vehicles, Stavens advanced perception for off-road autonomy. He developed a **self-supervised monocular road detection system** (333 citations) that could adapt to changing desert terrain, and a **self-supervised terrain roughness estimator** (39 citations) for predicting vehicle shock. He also explored **sub-meter indoor localization** using inexpensive sensors (57 citations) and **unsupervised learning of invariant features from video** (42 citations). His work on **online speed adaptation** (26 citations) addressed the critical but often overlooked problem of choosing the optimal speed for high-speed off-road driving. Stavens’ research remains a cornerstone of field robotics, demonstrating how robust perception and learning can enable autonomous operation in the world’s most challenging environments.
Research Focus
Key Achievements
Top Papers
- 1Stanley: The robot that won the DARPA Grand Challenge2,109 citations · 2006
- 2Junior: The Stanford entry in the Urban Challenge1,004 citations · 2008
- 3Self-supervised Monocular Road Detection in Desert Terrain333 citations · 2006
- 4Stanley: The Robot That Won the DARPA Grand Challenge234 citations · 2007
- 5Junior: The Stanford Entry in the Urban Challenge144 citations · 2009
- 6Winning the DARPA grand challenge with an AI robot86 citations · 2006
- 7
- 8Unsupervised learning of invariant features using video42 citations · 2010
- 9
- 10