We present a robotic table tennis platform that achieves a variety of hit styles and ball-spins with high precision, power, and consistency. This is enabled by a custom lightweight, high-torque, low rotor inertia, five degree-of-freedom arm capable of high acceleration. To generate swing trajectories, we formulate an optimal control problem (OCP) that constrains the state of the paddle at the time of the strike. The terminal position is given by a predicted ball trajectory, and the terminal orientation and velocity of the paddle are chosen to match various possible styles of hits: loops (topspin), drives (flat), and chops (backspin). Finally, we construct a fixed-horizon model predictive controller (MPC) around this OCP to allow the hardware to quickly react to changes in the predicted ball trajectory. We validate on hardware that the system is capable of hitting balls with an average exit velocity of$11 \mathrm{m} / \mathrm{s}$at an 88% success rate across the three swing types.

LiDAR is a method commonly used for measuring distances by illuminating the target with laser light and measuring the reflection with a sensor. Information provided by these scans has the potential to assist with disaster recovery efforts as they can help to survey infrastructure even when physical access may be limited. This research paper explores the feasibility of using a Random Forest classifier on limited hand-labeled data to identify buildings, along with a two-step filtering process that leverages unsupervised learning. Specifically, this two-step process uses a radius KNN filter followed by agglomerative clustering to identify and remove false-positive points. Results show the Precision of the classification algorithm increases greatly from 0.23 to 0.62 corresponding to a reduction in the false-positive rate which comes at the expense of some performance in the Recall but with the added benefit of building extraction.