Formula 1 talks up use of AWS AI for TV in 2019

Formula 1 managing director Ross Brawn explains how the sport will use machine learning to improve engagement

Formula 1 is to leverage three new AI integrations for its TV broadcasting in 2019 in a bid to better engage with viewers, and is using AWS to facilitate the move.

Managing director of motor sports at Formula 1 Ross Brawn told delegates at the keynote at AWS Re:Invent in Las Vegas this week that the organisation chose AWS to help unlock a lot of the data in what he called "the most data-rich sport in the world".

Each Formula 1 car has over 120 sensors, and more than 1.1 million data points transmitted between car and pits in a race. Each F1 team has its own CIO and IT team that use their own sophisticated analytics platforms, cloud services and machine learning algorithms, and now the organisation facilitating the sport wants to ramp up its use of data to improve the races and the engagement with fans.

As part of this, Formula 1 is using AWS high-performance computing (HPC) to develop better and more raceable cars, and machine learning models and AWS Sagemaker to make predictions of what can happen during a race.

Dubbed F1 Insights, the organisation will introduce this extra information for viewers on their screen next year.

"We can know if a driver is in trouble, if their rear tyres are overheating; we can look at the past performance of the car and where he is in a race, and machine learning can understand it and give that information to fans," Brawn said.

Broadcasters will also be able to show whether a driver is likely to overtake another driver in real-time; again basing this on past performances, and a range of other data that will be built into machine-learning models.

One area where Formula 1 will be focusing on that the separate F1 teams don't already have data on is on the tracks themselves, and where it can create tracks that enable better racing and more overtaking. Ultimately, this will lead to better engagement from fans, the organisation hopes.

"Can we change the format to make it more exciting and create models for that?" Brawn questioned.

He suggested that this could include looking at how the starting grid could be altered.

Brawn emphasised that these are all questions that Formula One can start answering now thanks to AWS and machine learning.