 Solar power is great, but the problem is it's an uncontrollable power source. It just creates a lot more variability and makes maintaining an efficient grid much more problematic. So I'm trying to develop methods to predict how much solar power is going to be on the grid. I've built a sky imaging system that is specifically designed to forecast solar power. It is off the shelf, 360 camera, an irradiance sensor, temperature sensors, humidity sensors, an onboard computer, a microcontroller to control the system and take data. It has onboard comms. We have a version that has PV power and batteries so it can be fully wireless. That's basically designed to be able to deploy it anywhere, flexible and robust and inexpensive.