I caught a presentation from John Boss of WeatherBug at the GITA event in Phoenix last week. The company is doing work with an IBM Research project titled Deep Thunder for high-resolution weather predictions for businesses with weather-sensitive operations, and one of their clients is the large utility Consolidated Edison.
The modeling project provides a very local forecast at 24-48 out, and has been in operation for a year. The highly local forecast factors in localized effects and has proven to be helpful in forecasting outages and allocating field crews. The success of the project has ConEdison committed to the project for the next five years.
With the tunnel vision on the future of sensor webs, it’s easy to forget that there are existing networks of sensors now that are providing new levels of insight into business operations. It’s great to hear of such a practical example that is yielding strong business results.