In today’s power grid management precise prediction are needed as the variety of production methods and consumption changes fast. There’s a growing need to predict the production to enhance the efficiency of the system. Traditional power production is easy to adjust according to the needs of the grid, but many renewable sources are harder to predict. In this project we are interested in perfecting the photovoltaic (i.e. solar) power prediction. Currently there are ways to predict the daily production of solar panels, but could there be way to make this into 30 minute or even shorter cycles? There are few critical aspects that define the power production, such as time of day and cloud cover.​ How might we utilize accessible and cheap camera technologies to scan the sky in specific locations to determine the amount of short-term solar radiation? How might we be able to measure cloud cover and how it is moving? What kind of set-up would be needed to calculate the predictions on the spot (without the need to use cloud computing)?​ In this project we are going to physically build and test the prototypes in real-life environment. Join the team and put your theoretical knowledge in practice!