Mohammed Kabir

I'm a roboticist and I specialize in vision-based GPS-denied guidance systems for aerial robots and spacecraft. I have a keen interest in vision-based navigation, planetary EDL (entry, descent and landing), and high bandwidth datalinks. I have over a decade of combined experience working at the intersection of industry and academia, in the fields of computer vision, state estimation and automatic control. I currently lead engineering at Corvus Robotics (YCombinator S18), where we're solving some of the toughest problems in lifelong autonomy for aerial robots. Our robots, deployed in industrial environments, push the state-of-the-art in robust visual-inertial odometry, long-term visual relocalization, semantic scene understanding and realtime motion planning. Outside my usual roles, I've been a long term contributor to open-source software projects (PX4, ROS) and enjoy interdisciplinary work in architecting complex hardware-software systems and designing novel sensor hardware. In my free time, I enjoy photography and dystopian fiction.