IBM's Watt-sun: great at solar forecasts, useless at Jeopardy
Pop quiz: if you worked on a research project at IBM that dealt with chewing on large amounts of data to predict how efficient solar power arrays could be, what would you call it? If you said "Watt-sun," you'd be on a roll. Tongues may have been planted firmly in cheeks when the name suggestions started flying, but Watt-sun is no joke. It's a solar forecasting platform that continually mashes up data from scads of sources -- from existing models to satellite views to cloud cover imagery captured by cameras lashed to poles -- to try and predict much sunlight solar panels will actually be able to suck up.
Why does that matter? Well, if you were a power company fleshing out your part of the power grid with solar cells, you'd want to know how much juice to expect from them. Factors like weather and the intricacies of the sun's path across the sky make forecasting tricky, but Watt-sun project manager Hendrik Hamann told Gigaom that Watt-sun is about 35% better at its predictions than the next best thing. Watt-sun's been tested in about a dozen solar sites around the country and if it's a good as its creators say, it could be a big help for a country struggling to wean itself off dirtier forms of energy. Alas, the team behind the project doesn't plan to turn Watt-sun into an actual product -- here's hoping they change their minds.
[Image credit: Flickr/Michael Mees]
