Situation
Outages are fairly common in power distribution networks and this number is increasing in some countries because of aging infrastructure and changing weather patterns. Power outages manifest themselves in multiple ways. Lost revenue to the utility. The expense to repair the damaged assets to restore power. Inconvenience and lost productivity to businesses and consumers. Possibility of fines and penalties. Reducing the number of unplanned outages and better managing their duration is a priority for most utilities to mitigate costs and minimize inconvenience to the customers and possibly save lives.
Solution
The Utilihive platform provides access to data relating to weather, IoT sensing, smart meter and geospatial technology allowing utility dispatcher to have access to more and better data to realistically predict when and where outages are most likely to occur. Modeling outage risk requires AI to quickly ingest and accurately analyze multiple geospatial datasets. Datasets such as wind speed, wind gusts, tree canopy, population density, snow and icing, lidar and vegetation information and a utility's historical asset information and observed outages are essential to calculate the timing and likely location of outages. What’s more, machine learning can be built into the modeling process to continually improve the predictive analytics.
Benefits
Better predictions of outages can significantly reduce their impact or even prevent them. Having advance knowledge of where outages are most likely to occur allows companies to position crews and resources close to those places so they can restore power more quickly. It also helps them ensure they have enough resources to restore power after a large-scale incident. Having the insights as to which areas are at risk due to old infrastructure or more vegetation may even enable utilities to prevent some outages entirely.
