Improved the NREL Panel-Segmentation package on GitHub by training a convolutional neural net to detect solar-arrays from satellite images. Resolved issues regarding overlapping of detection boxes as well as smoothly integrated the CNN into the Panel-Segmentation Package.

Worked on integrating a cost model for battery units into a larger cost model for hybrid power plants specifically including solar-arrays and wind turbines. On top of this integration, I included the capabilities to include different type of cable systems and to give each module a coordinate and calculate the model based off of real measurements rather than an average setup. This can be all found in the HybridBOSSE package on GitHub.

Taught kids how to program and walked through with them to debug their code.