Noise to Knowledge

An innovative company specializing in traffic infrastructure wanted to transform raw data from their equipment into a real-time streaming API for connected and autonomous vehicles. The data volumes are a tremendous challenge, as each intersection can generate more than 40 GB per day, and a small city’s traffic infrastructure will generate more data than Twitter on a daily basis. With more than 300,000 intersections in the U.S. alone, building a traditional software architecture that scales to millions of vehicles is not cost effective. Even more, the latency incurred is unacceptable as drivers and autonomous vehicles make decisions on a sub-second basis.

<300ms

end-to-end latency

110

neural networks

95%

data reduction

The traffic company used Swim to successfully launch their new service with unmatched performance and scalability. The API now transmits full-resolution data with a 100-200 millisecond latency at a fraction of the cost from traditional architectures. Each sensor with a real-time data feed was modeled with Swim, to include traffic controllers, crosswalk signals, vehicle detectors, and then deployed on the supporting transportation agency’s infrastructure. Swim dynamically generated a stateful replica of the entire real-time data stream and reduced network traffic by more than 95%.


Traffic in The Now

Each sensor with a real-time data feed was modeled with Swim, to include traffic controllers, crosswalk signals, vehicle detectors, and then deployed on the supporting transportation agency’s infrastructure. Swim dynamically generated a stateful replica of the entire real-time data stream and reduced network traffic by more than 99.25%.


Other Streaming Apps

Want to see how we’ve used Swim to build other streaming applications? Check out some of our other projects to experience Swim in action and learn tips, tricks and best practices for building streaming applications.


Swim applications continuously provide the latest updates to empower users with the most relevant insights. Swim doesn’t just stream data; it runs application logic in-stream to power real-time analytics, business workflows, dashboards, and UIs. Instead of out-of-date snapshots, Swim mirrors your entire application in real-time.