The issues we chose solve was pertaining to renewables. Due to the large fluctuations in the power that can be generated from wind turbines, making use of them to power the grid becomes a complicated affair. When there is low power generated, the power station has to be able to anticipate and make up for it. We also learnt about more about the nature of wind turbines, where wind speeds exceeding the optimal range also had to be cut off, so as to prevent the turbine electrical generator from damaging. As for the data set, we sourced it from an international weather website

We were then able to improve upon the use case provided and propose areas where building the turbines would be optimal, on top of predicting the speed. For the prediction, I used the past data for the wind speeds in the area, and build multiple models, such as Ransac regression and using multiple variables to achieve the best result.

Throughout the course, we attended full length courses on the predix framework, which usually cost thousdands of dollars, for free. There, we got the opportunity to skype conference with industry experts, and learnt more about the predix software.

In the end, we came out on top as winners in the renewable Category, being able to bring back a cash prize of $300