Clean vehicle traces with the Map Matching API

Since launching the Map Matching API, we’ve seen a lot of great examples that turn noisy traces into usable routes. While there are many creative ways to utilize this API, one of the most powerful and valuable applications is the output of a clean route from raw vehicle traces.

Even the best location detection technology has some degree of imprecision with each capture. Satellite-based systems and many others rely upon trilateration to determine distance from known points, and any indirect paths the signals take introduce errors. While raw traces have immense value by themselves, in their original state they almost always fail to provide an accurate representation of distance or geometry traveled on a road network.

Map Matching in action at Vinli

Our friends at Vinli have put the Map Matching API to work, allowing developers to visualize completed trips as clean routes. Their OBD-II device turns almost any car into a connected car by regularly streaming a read-only feed of vehicle diagnostic data and location information to Vinli’s platform. While there are many ways to view this data (APIs, SDKs, CLI, etc), Vinli’s Developer Portal provides an easy way to visualize telemetry, trips, and much more. The Developer Dashboard displays raw geographic data under Locations, which strings together original telemetry messages as interactive points into a single trace.

Go to Trips on the same dashboard, and you will see something slightly different. While the raw locations are available in near-realtime, Vinli’s platform uses the Map Matching API upon completion of a trip to provide a cleaned up overview of the whole journey.

Finally, here's a simple animation to highlight the differences between the raw trace and the resulting clean route.

Get started

To bring clean traces to your own application, first check out the new Map Matching API Guide. The full API documentation will help you across the finish line.