Map Matching

We offer a high performance map matching library for both embedded devices (thick clients) and well as server based (thin clients) solutions.

Find out more

REQUEST A CALL BACK

close

What is Map Matching?

Our algorithms and software modules have been used and approved in many field tests and industrial projects, e.g. in the German road pricing project.

Any global satellite positioning system (GNSS) like GPS, GALILEO or GLONASS , which is used for real-time positioning includes inevitable errors. Additionally, there's no (and will be no) absolutely accurate digital street map representing the real world. Therefore, in order to show a vehicle's position on a digital street map you have to match ("snap") its coordinates (received from a global positioning system) on top of a vector based street map. This challenging process is called "Map Matching".

Utilizing our UMDB and MicroMap frameworks we can offer a powerful map matching software that runs virtually on any vector-based map. We do not rely on map providers like Navteq, Tele Atlas, AND, or other commercial map vendors. We even support open source maps like OSM.

Personal navigation devices use on-board map matching algorithms to calculate the "true" vehicle position using a digital vector-based street map. The integration of additional hardware sensors makes a positioning device more expensive and prohibits simple turn-key products. But even with very good sensors, an initial sensor calibration and continuous sensor fusion these inaccuracies can't be avoided.

Smart Map Matching

Since both the user's location and the underlying road network is not 100% accurate, you have to snap the location obtained from e.g. a GPS receiver to the nearest node or edge in the vector map. In general, we know Point-to-Point, Point-to-Curve and Curve-to-Curve map matching algorithms. Of course, today these basic methods are improved and refined by considering rich map attributes like road class, turn restrictions, lane information, etc.

We've significantly improved the quality of our map matching algorithm by applying a configurable vehicle model which reflects the physical characteristics of a vehicle.

Our map matching algorithm continuously reads incoming GPS data and checks each position information whether it complies with the underlying physical vehicle model.

Sparse GPS positions & Road Pricing

In case of missing GPS signals or unreliable position information we have to deal with sparse trace points. In order to reconstruct the vehicle's real trajectory we have to consider and evaluate all possible paths and directions the vehicle might have taken. This is done best with routing algorithms which we've integrated in our map matching algorithm as well.

An on-board GNSS devices (e.g. GPS) provide a continuous data stream of real-time position information. In most cases these positions have to linked with a digital road or street network in order to show the vehicle's current location on a map. Various map matching algorithms are in place to snap disturbed positioning data (caused by signal multipath problems) to a set of road segments. Satellite-based road-use charging systems need to perform this task absolutely reliable and of course very efficiently.

Vehicle Model

The vehicle model requires a minimum set of parameters initially passed to the map matching process:

  • the vehicle's maximum speed
  • the average GPS accuracy
  • vehicle's maximum angular velocity
  • realistic acceleration and deceleration coefficients
  • general static friction coefficient

Found in these Solutions and Products

Nav SDK

Easy way to get own navigation software

Geo SDK

Easy to use geocoding service

Map SDK

Different map rendering features

Web Server

Our solutions in your cloud

In-Dash

Extended capabilities on in-dash devices

On-Dash

CloudMade enables a range of On-Dash devices

Contact Us to Get More Information

Juha Christensen

CEO

juha@cloudmade.com

James Brown

CTO

jim@cloudmade.com

Nick Black

VP Product

nick@cloudmade.com