Predictive Analytics - Informing Disaster Preparation and Response

 A longtime presence in the NYC entrepreneurial community, Roger Coleman leads Disaster Tech and focuses on pathways of identifying dangers and preventing risks associated with natural disasters. One area of focus for Roger Coleman and his New York team is predictive analytics that help assess energy infrastructure risks and vulnerabilities.

This type of analysis relies on existing data sets in identifying patterns and quantifying the likelihood and potential severity of a disaster. Big data may be applied to a specific location, as when a municipality in a valley evaluates the risks of a major landslide, given weather, geology, vegetation, and other variables. In other cases such analytics can span vast areas of the globe, such as when assessing the risks of hurricane.

The value of big data is in the sheer amount of information that past natural disasters have provided. For example, every time a storm hits, sensors generate real time data on weather patterns, wind levels, and rainfall. Early warning signs that make it likely that Category 1 rainfall will develop into a Category 5 storm can be extrapolated from the thousands of storms on record.

At the same time, machine learning algorithms enable data to be automatically collected and analyzed, with each additional set of data points providing deeper and more accurate understanding. The end result is that local and national agencies are able to set up relief protocol appropriately sized to the actual level of risk and its local impact.