Ninja Car. Simulation-in-the-loop planning and control. Rigorous justification for model-predictive control with high-order nonlinear plant models and optimization. Real-time perception algorithms for localization with various mapping representations (e.g. sign distance function fusion). Watch a video of the hydraulically-actuated car jump!


Camera calibration tools. There are a bunch of ARPG utilities that deal with supporting perception development. One of the tools to which I have contributed the most is Calibu, a library that calibrates arbitrary rigs of cameras and IMUs for production by estimating extrinsics (e.g. positions of sensors) as well as intrinsics (biases, generic parameters).

Previous research projects were:

Theoretical swarm dynamics modeling in robotics and distributed control algorithms with my collaborators at NRL and Drexel.

Autonomous boat control and probability-based control of underactuated vehicles (NRL/Drexel collaboration).

Slow-fast systems and dimensionality reduction for predicting rare events of stochastic systems operating on multiple timescales (NRL).

Asymptotic analysis, bifurcation and control of delay-differential equations of coupled Rayleigh-Plesset Equations (Cornell).