Analysing GPS probe data will unlock driver services

The concept of the connected car no longer refers to a near future scenario – it is already here. More than a million drivers already share GPS position data in real-time during most of their drives, generating a data stream of several billion GPS probes per year. This so-called ‘Big Data’ can be harnessed in … Continued

The concept of the connected car no longer refers to a near future scenario – it is already here. More than a million drivers already share GPS position data in real-time during most of their drives, generating a data stream of several billion GPS probes per year. This so-called ‘Big Data’ can be harnessed in numerous ways to create innovative services that make the road a safer, greener and more pleasant place.

Real-time traffic is the most obvious use case of GPS probes. Companies such as Nokia, Inrix and TomTom, for example, have been using this data as a cost-effective source to power their traffic services in recent years.

More than a million drivers already share GPS position data in real-time during most of their drives, thus generating several billion GPS probes per year

Moving forward, as more car electronics get connected to the cloud, car data from the CAN bus could be exposed and used in innovative services. For example, data from windshield wipers or ABS can complement GPS probes to inform other drivers of severe weather conditions, slippery roads, etc. Global positioning data and an accelerometer can also warn drivers in real-time of emergency braking events happening ahead of them on motorways.

But all of this, and more, is only possible if the GPS probe data is accurate, and if the cost of transferring this data to the cloud is affordable.

A probe is usually made of a string of data: latitude, longitude, altitude, heading, speed and, most importantly, precision. The latter can vary considerably, being influenced by many factors, including the quality of the GPS chipset and its antenna; vehicle modifications that may jam GPS signals, such as heat reflective windshields; and the location of the vehicle – urban canyons, tunnels, etc. are well known to obstruct GPS and create false data. To produce a clean dataset, GPS readings coming, for example, from parked vehicles should also be filtered.

The real innovation lies as much in the data the connected car creates as it does in the information it receives

Fortunately, GPS inaccuracies can be mitigated on the device by additional sensors such as a gyroscope, accelerometer, pressure sensor (for altitude) and tachometer, to name but a few – and reduced by techniques, like map matching, which eliminate or rectify inaccurate data not located on the road network.

Moving forward, precision will also be improved through additional sensors such as radar, cameras and new GPS-like satellite constellations (Glonass today and Galileo tomorrow).

Meter-level precision will create another wave of innovation. One could think of a traffic service that would recommend the fastest lane on a packed motorway, based on lane-by-lane real-time GPS probe analysis.

The decreasing cost of M2M wireless data will also allow cars to exchange more information, more frequently and unlock services and business models which are not sustainable today.

In a nutshell, the real innovation lies as much in the data the connected car creates as it does in the information it receives.

The opinions expressed here are those of the author and do not necessarily reflect the positions of Automotive World Ltd.

Jean-Marc van Laethem is Co-founder and Chief Technology Officer, Coyote System

The AutomotiveWorld.com Expert Opinion column is open to automotive industry decision makers and influencers. If you would like to contribute an Expert Opinion piece, please contact editorial@automotiveworld.com.

Welcome back , to continue browsing the site, please click here