Innovation: A GNSS Odometer

April 1, 2008 By: Andreas Wieser GPS World

How Far Have We Come?

INNOVATION INSIGHTS With Richard Langley

WHAT DO THE GREEK MATHEMATICIAN Archimedes of Syracuse, the American statesman and polymath Benjamin Franklin, and the Mormon pioneer William Clayton all have in common? They each invented an odometer — a mechanical device for measuring distance. Whether we be military engineers, mail-route mappers, wagon masters, or just automobile drivers, we often want to know not just where we are but how far we have come.

Richard Langley

The odometer was likely first invented by Archimides during the First Punic War when Syracuse got in the way of Rome during its battle with Carthage. A Greek origin is fitting as the word odometer derives from the Greek words hodós, meaning "path" or "way" and métron, meaning "measure."

The device was reinvented many times over the years but its use was not widespread until the development of the automobile, and now virtually every vehicle sports one. Mechanical odometers gave way to electronic ones but the distance traveled was and is still determined by counting wheel revolutions. But just how accurate are the odometers in our modern vehicles? Not very, it seems. The odometer reading is affected by tire pressure, tire slip, and incorrect calibration. And while in many countries there is no regulation covering odometer accuracy, the Society of Automotive Engineers' voluntary standard and that of the European Commission is only plus or minus 4 percent, or as much as a 4-kilometer error in every 100 kilometers.

Does this matter? Well, in effort to reduce the cost to the general tax payer of maintaining roads or reducing conjestion, many administrations have implemented "road pricing," where a flat charge is levied for using a particular stretch of road or for entering a city center. But some administrations are charging per kilometer of travel with data coming from an odometer recording. Automobile insurance companies have also implemented plans where the premium is based on the distance traveled by the vehicle ("pay as you drive"). To fairly implement such schemes, governments should require more accurate odometers in vehicles. Could an odometer based on GNSS be a solution?

In this month's column, we take a look at how distance traveled can be computed from GNSS observations and just how accurate those computations are.

"Innovation" is a regular column that features discussions about recent advances in GPS technology and its applications as well as the fundamentals of GPS positioning. The column is coordinated by Richard Langley of the Department of Geodesy and Geomatics Engineering at the University of New Brunswick, who welcomes your comments and topic ideas. To contact him, see the "Contributing Editors" section.

The possibility of replacing a traditional odometer by a satellite-navigation-based system naturally allows combining information on the distance traveled with information on location and time. These are pieces of information increasingly requested by road authorities, insurance companies, and company executives to monitor driving patterns and reduce non-commerical traffic or distribute it better spatially and temporally. This is and will be accomplished mainly by financial incentives such as road pricing and pay-as-you-drive insurance.

A "GNSS odometer" — a stand-alone GNSS receiver with dedicated software for computing the distance traveled — can be economically manufactured as a small, light-weight, and power-efficient unit. The required accuracy of the odometer within a tachograph is 4 percent as specified by the European Commission. If this accuracy can be achieved without integration of other sensors, and without need for map data, a GNSS odometer would offer significant benefits over conventional odometers: it has no mechanical parts (and thus no wear); allows for easy installation in both new and used vehicles; is not affected by tire pressure, vehicle load, and road surface conditions; and is tamper-proof.

Some researchers have mentioned the possibility of computing the distance traveled by a vehicle from the consecutive position solutions output by a GNSS receiver; results suggest that an accuracy of a few percent is attainable. Using the coordinates for distance computation seems a natural choice, especially if the computation is to be accomplished within a navigation system where the results can possibly be improved by using available map data. However, in view of transnational traffic, cost, and potential applications beyond road-based traffic, it is preferable to find a solution that does not require map data.

An approach with significantly higher attainable accuracy is briefly summarized herein, using Doppler-based velocity estimates rather than pseudorange-based coordinates. Further details are available in the comprehensive report (see Further Reading) and the research paper on which this article is based.