Architecting a control system for autonomous bulldozing

E. Harber, E. Schindewolf, V. Webster-Wood, H. Choset, and L. Li, “A Tunable Magnet-based Tactile Sensor Framework,” IEEE SENSORS Conference, October 2020. (forthcoming)

Developing architectures for a wide range of sensing applications achieving solutions for < 1 to 30 kg robots

Testing and analyzing elastomers materials and structures, and subsensor locations to minimize hysteresis and environmental effects increasing signal-to-noise ratio by at least an order of magnitude

Preparing data, analysis, and visualizations for demonstration and patent application

Implementing a custom cost function for use in multivariate gradient descent

Integrating structural decision making for external effect cancellation

Utilizing parallel processing and vectorization for > 20x speed increase over otherwise optimized code

Drove development of a novel pushing controller from problem statement to physical testing and achieved

Established a Gazebo simulation environment for virtual testing from scratch