Using machine learning to predict the risk of sewer backup
The Insurance Bureau of Canada tasked Tesera Systems with creating a web app to help analyze the risk of residential sanitary sewer backup to municipal infrastructure.
Tesera built their solution using FME Flow Hosted as the data processing infrastructure, Amazon Web Services (AWS) S3 for data storage, and AWS SQS for task queuing. They set up FME to watch for SQS messages, process GIS data submitted via the app, perform data validation, and output model indicators for further analysis. They performed machine learning on the resulting datasets, which integrated insurance claims with infrastructure data to help predict the risk of a sewer backup.
Minimizing costs and maximizing value
Using FME Flow Hosted for this AWS-based project enabled Tesera to automatically process huge volumes of data in the cloud while keeping infrastructure costs low. Since FME Flow Hosted is hosted in AWS, they were able to leverage data gravity and create a performant solution. Municipalities using this web application can take action to improve infrastructure in high-risk areas and identify which areas require flooding emergency plans, resulting in valuable risk mitigation and cost savings for the Insurance Bureau of Canada.
Tesera Systems is a Canadian consulting services company that integrates FME into its innovative data-driven solutions and geospatial applications.
