[RACING] GPS accuracy in the EQUIMETRE solution


Blog GPS copie

In training monitoring solutions, GPS is a key element. Indeed, it is responsible for the tracing of the workout, the distance covered and above all the speed. When training a racehorse, the calculation of speed must be extremely accurate in order to provide the trainer with reliable data to assist him in his decision making.

On the Equimetre platform, three elements use the GPS: the training map, the key speed parameters, and the intermediate time table. Each of these tools use the performance of the Equimetre GPS to provide accurate and objective data that complements the feeling of the trainer and the rider.

The accuracy of the GPS is therefore essential, which is why the Research & Development teams at Arioneo pay so much attention to it.

Why is it critical to have accurate GPS data and how can the trainer use it? 

  • Playback of trainings from the satellite map: analyse the speed at a given point on the map.
  • Playback of trainings: speed and distance at any time of the training
  • Key parameters for each training and longitudinal analysis: Max speed, Best 200m, Best 600m, Best 1000m
  • Objectification of intuitive feelings
Screen Shot 2020-04-10 at 17.28.39
Capture d’écran 2020-05-04 à 11.56.33

How can split time analysis assist trainers?

  • Analysis of interval training: adjustable interval distances
  • Kilometric reductions
  • Acceleration analysis in relation to physiological and locomotor parameters: how did my horse cope with the speed work?
  • Analysis of speed in relation to topology and cardio: was the ascent the most difficult part for the horse?
  • Decision making: comparing training data with race split times to plan a race.


Guillaume Dubois, PhD, Scientific Director, and Clément Beaudoing, Software Engineer, both members of Arioneo’s Research and Development team, tell us more about the accuracy of EQUIMETRE‘s GPS.

  • How does the GPS work ?

In space, there are many constellations of satellites that orbit around the earth.  A “constellation” is a group of satellites which is used to define a position.  There are several types of systems: the American system GPS (Global Positioning System), the Russian system GLONASS, the European system Galileo, which is starting to be used, and finally the Chinese system Beidou, which is currently being deployed.

Each of these systems is made up of around twenty satellites which orbit around the earth. By connecting to and combining these different constellations we can increase the accuracy of our GPS.  The satellites send a signal, indicating the time. Our GPS unit receives these signals and combines them to estimate its position in relation to the position of the satellites in space.

  • What does the accuracy of a GPS correspond to in practice?

A GPS’s accuracy can be ascertained by measuring the gap between the position indicated by the GPS and its actual position.  A GPS can generate two types of errors.  First, on a route, a recurring “lag” error can emerge.  For example, all of the measurements are slightly off the actual position by ten or so centimeters to the right. 

Second, there are more subtle errors.  Putting to one side the “lagging” type of error (described above), we can nonetheless encounter errors linked to the GPS’s internal accuracy.  These errors may be linked to the atmospheric conditions or to reflections (i.e., signals bouncing off something in the environment).  For example, if you are close to buildings, the signal emitted by the satellite will bounce off the buildings and back to the GPS.

The additional time taken for the signal to travel through the atmosphere will distort how the GPS interprets the satellite’s position in space.  In addition, if the sky is very cloudy, the signals will be slowed down as they move through the atmosphere. This means that the sensor will indicate that the satellites are further away than they actually are.

  • What are the different factors that impact a GPS’s accuracy?

To improve a GPS’s accuracy, we can combine different frequency bands.  There are also correction methods that can be used, such as installing a fixed ground unit, the position of which we know perfectly.  This gives us an idea of the margin of error that exists between GPS’s measured position and its actual, correct position. We can use the data that we obtain with EQUIMETRE to identify such errors and improve the accuracy of our GPS.

Finally, if the GPS is partially covered, for example underneath a horse, this may lead to errors in the GPS data.

  • The GPS has a frequency of 1Hz. What does that mean?

EQUIMETRE’S GPS has a frequency of 1Hz, which means that it obtains one GPS position per second.  In other words, every second, the GPS sends us its position.

  • Does the 9 axis/gyroscope sensor affect this accuracy?

The 9-axis sensor does not directly affect the accuracy of the GPS, but it can be used to make corrections. As it is an inertial unit, it estimates position, speed and acceleration at a specific point in time.  If the inertial unit indicates that the horse is going straight for 10 seconds, while the GPS indicates that it turned off-course for 10 seconds, one of the two must be incorrect. It is thus possible to confirm or correct the position given by the GPS using the data from the inertial unit.

  • What were the biggest difficulties that you encountered in achieving this level of accuracy?  What were the different stages of the development process? Did you use existing, off-the-shelf technology?

One of the biggest difficulties we faced when integrating GPS into EQUIMETRE was working with EPOs (Extending Position Orbit).  EPOs indicate the approximate position of satellites over the course of a fortnight. They allow us to receive GPS signals faster, which in turn allows us to identify a position faster and more accurately. If the signal is lost for a few seconds, with EPOs we can obtain the GPS position much more accurately.  On the other hand, EPOs have to be updated in EQUIMETRE every two weeks, which means additional Bluetooth exchanges.  

First all we looked for a sufficiently accurate product, with an integrated antenna, which was also both energy efficient and compact, and thus could be easily integrated into EQUIMETRE.  We had to find a GPS satisfying all of these criteria amongst the vast range of products existing on the market. It also had to exist on a development map so that we could test it easily.

As regards development, the implementation of EPOs took longer than we anticipated because the exchange of files is quite significant.  Next we had to create library to download the information and then we had to work on configuring the system.  We had the choice of various options, such as: using an assisted GPS or using additional information to achieve greater accuracy, using GPS alone or combining GPS, Glonass and Galiloe, etc.

  • How can you use GPS to measure speed ? 

There are two methods for measuring speed using GPS.

The first method is to measure the time elapsed between two successive GPS positions. We can then measure the distance between these two positions

  • If a GPS is accurate as to location does this automatically mean that it is accurate as to speed?

The accuracy of a GPS doesn’t automatically mean that it will give accurate results as to speed.  These are two different types of calculation, which can thus lead to different results.

  • What is the benefit of greater accuracy for the user?

The benefit of greater accuracy depends on the type of measurement that the user is looking for.

If the speed needs to be 0,01 km/h, then the GPS needs to be very accurate.  If the user wants to know if the horse is between 63 or 64km/h, then accuracy is less important.  For example, trainers often look at the time that it takes for a horse to run 200m.

With a GPS with 5m accuracy, the margin of error when measuring time over 200m is around 3.5%, or a tenth of a second and it is sufficient most of the times.



EQUIMETRE is a technology dedicated to the race horse training and allows to measure :




Evolution cardio cheval entrainement