• We identify and assemble the optical data for the area the customer wants.
• We select two of the most recent data sets available for the region.
• The timeframe of the data availability will vary state to state and we can provide the customer with a more accurate timeframe for their state.
• The model compares the two datasets to establish what has changed.
• We have trained the model to identify where there are data clusters of change that correspond to a house.
• Examining the data clusters, the model identifies which are likely to be houses based on the parameters we have given it.
• We use machine learning to improve the model using ground truth data, so it can distinguish what is a house in that particular region versus what is perhaps shadowing, a roundabout or a pond.
• We upload a shapefile with the inundation zones, we can start with using FEMA’s Special Flood Hazard Area (SFHA) data, which delineates the area that would be inundated by a so-called 1 in 100 year flood event and is used to enforce where Flood insurance is compulsory.
• The product highlights where new buildings intersect with the flood area.
• Within the three year subscription, we will provide an additional dataset when it comes available along with the change detection analysis.