How do we recognize a real innovation? The Massachusetts Institute of Technology (MIT) tried to answer this difficult question with the publication of the following article in its Technology Review: “The ten breakthrough technologies of 2014”. More than allowing a dramatic change in our day-to-day habits, the 10 laureates have another common point: thanks to them, some of the most difficult scientific problems will be able to be solved. Thus, we can say that those technologies serve global progress. Supporting medicine – with brain mapping – as well as freedom of speech – thanks to the secure phone -, each of them represent, according to the MIT, a potential revolution.
Forecasting and renewable energies brought to another level
Can forecasting really be revolutionary? It’s true, using data – like satellite images – in order to predict meteorological evolution is not a new thing. But the difference today is that we can collect an increasing number of data. Wind turbines can for instance transmit in real time a huge amount of indications: generated power, wind speed and direction, turbine speed…
At the same time, very precise statistical models that can translate those data into predictions have been developed. Associated with an increasing number of information, this phenomenon enables to manage very precisely wind power, thanks to accurate forecasts. Indeed, the margin of error in wind forecasting used to be not far from 20 percent while it can reach 5 percent or less today, depending on the time horizon considered.
As far as solar energy is concerned, the « forecasting revolution » follows the same tendency today than wind forecasting did in the past ten years. The amount of data collected keeps on increasing, coming simultaneously from meteorological imaging, satellite and ground cameras. Thus, we can now have in real time all the information that we need from each solar panel – irradiance, temperature, generated electricity… After analyzing and comparing them with what we know from the solar installation (orientation, geographic localization, past data), we can forecast potential photovoltaic production.