Automating Geospatial + Predictive Analytics to Accelerate Insights and Mission Outcomes
Advanced Geospatial Analytics
in Disaster Response
The proof of this comes from some robust geospatial related real-world uses cases developed by an Alteryx partner, Atkins, a contractor to the U.S. Federal Emergency Management Agency (FEMA). When dual hurricanes hit Puerto Rico in 2017, Atkins used the geospatial and predictive analytics capabilities found in the Alteryx APA Platform to perform automated substantial damage estimates for close to 150,000 structures, identifying and prioritizing those structures most in need of a physical evaluation.
The strategy was to use Alteryx to estimate the structural damage that had occurred, prioritize areas that still needed some sort of human inspection, and ultimately reduce the total number of in-person inspections so that the recovery process could begin quickly. Alteryx was used to blend over a dozen datasets and the following variables to predict damage:
Atkins determined that they needed to get three functional groups of teams out to the field quickly. The first team used a geographic information system (GIS) to find the locations of damaged structures. The second team collected the information on the structures, and the third team built the analytics model in Alteryx.
The data that was feeding the model came from very disparate sources, including data from the European Union, NOAA, the National Weather Service, FEMA, and the Army Corps of Engineers. This caused some significant data challenges, such as overlapping building stock data, numerous databases of different ages and sources, an inconsistent addressing system, and multiple sources of inundation and flood depth data. Additionally, this data was stored in various formats that required cleansing prior to blending, and creation of indices between the datasets. Alteryx was used for these data prep and blend tasks.
At the heart of this work was the ability to use the Alteryx APA Platform’s predictive modeling capabilities to overcome some of the incomplete or missing data, by filling in the gaps. Additionally, the team leveraged a Boosted Regression Model to look for patterns in the data across thousands of iterations, selecting the most relevant variables and providing a high degree of control over the entire predictive model.
With this model, 146,039 structures were evaluated and sorted into priority groups for actual physical inspection. From those structures loaded into the model, it was determined that just over 30,000 of them required the dispatch of inspectors. With this geospatial and predictive capability, inspection resources were deployed where they were most needed, saving tens of millions of dollars in inspection costs and most importantly — speeding up delivery of recovery support to people and communities.
Staying Ahead of the Flood of Data at FEMA
This same ability to ingest, prep, join, and analyze massive geospatial datasets and apply predictive analysis in a unified platform enabled Atkins to work with FEMA to analyze and identify flood risk across 90,000,000 structures, throughout all 50 states and six territories. The Alteryx workflow ingested and blended over 300 datasets, with 4000 internal layers. The workflow contains over 3600 automation building blocks and produced actionable insight to inform federal, state, and local authorities on the risk faced by individual structures across different types of flooding situations.
Want to hear more on the subject?
Watch our webinar on-demand: Automating Geospatial Analytics and Predictive Analytics with Alteryx, featuring Michael DePue, Principal Technical Professional, Atkins Global.