How one company is trying to transform the Tour de France with big data and machine learning

For someone who doesn’t watch the Tour de France, the nuances of this multi-stage bicycle race can be quite difficult to grasp. While an outside observer might note that one team is winning and that another is lagging, the why of the matter will be obscured. Dimension Data, the official technology partner for the Tour de France, has some ideas on how to make this data more friendly and accessible to a larger audience.

Dimension Data has been working with teams and the Amaury Sport Organisation (ASO), the event organizer, to integrate sensors into the bikes. The data generated from these sensors is used to create what is effectively a real-time map of the circuit, and a lot more besides.

Data acquisition

Primarily, the data being gathered is real-time location data. One would expect that location data to be generated via GPS. On the surface, this seems to be the simplest and most logical solution to the problem. Dimension Data, however, handles this rather differently.

Speaking to Scott Gibson, Group Executive, Digital Practice at Dimension Data, we discovered that Dimension Data doesn’t put GPS sensors on the bikes. Gibson tells us that the GPS latency is too high for accurate, real-time analysis of the rider and his position.

Dimension, in fact, mounts what can best be described as transponders under the seats of the bicycles. The vehicles (cars and motorbikes) that move with the participants house receivers and create a sort of “moving mesh network”, says Gibson. This location data is then relayed to a helicopter, a fixed wing aircraft, and then to computer servers and trackers located at the finish line.

While the method does seem complicated, Gibson says that they’ve managed to drastically cut down on the latency. The complex setup does mean that at the moment, the range of this mechanism is limited to about 3 km. Any rider outside of that 3 km window cannot be tracked.

The data gathered is relayed to a website and app where anyone can see the speed, position and other details of the racers. This is updated at one second intervals. This data is overlaid over pre-loaded routes.

Interestingly, teams also have direct access to this data via an API, allowing them to also track their team members in real-time.

Currently, Gibson says that Dimension Data will collect over three billion points of data. This includes weather data and detailed track information.

When asked whether Dimension Data would collect biometric information during the race, Gibson said that while his company can do it, the data is sensitive and needs to be private. Biometric information includes factors like heart-rate, muscle fatigue, breathing rates, etc.

Teams would not want rivals to have access to this data as it would give competition an unnecessary advantage.

Dimension has developed algorithms to predict the outcome of the races. These algorithms use the gathered data as well as historical data for the cyclist, information on the races he's good at, etc.

Gibson says that they can use this data to predict the outcome of a race with some degree of accuracy.

Right now, they don't have enough data to provide any startling revelations and the like about the sport, but that might come soon enough. With the data it has, Dimension already claims to have transformed the viewing experience for fans, who now have more information on the tactics and nuances of the sport.

Intel, working with the ICC, is already in the process of bringing big data to cricket with the likes of the bat sensor and other tracking technologies finding their way into the sport.

As Gibson explained, the success of any sport boils down to viewership, and Dimension Data working with the ASO are able to improve this aspect.

The technology is also helping bring better insights to the sport by introducing an element of statistical predictability.

Riding for a cause

As an aside, Gibson tells us that Team Dimension Data, a team that's participating in the Tour de France, is sponsored by Dimension Data and represents a charity called Qhubeka. The purpose of the charity is to provide bicycles to African youth. Greater mobility is key to the growth of any culture, after all.


Published Date: Jul 04, 2017 08:39 am | Updated Date: Jul 04, 2017 08:39 am