How Formula 1 uses AWS to leverage remote operations and production for the Abu Dhabi Young Driver Test
This blog post was co-authored by Ryan Kirk, Cloud Architect and Cloud & DevOps Team Manager, Formula 1.
As the Lead Cloud Architect and Team Manager at Formula 1, I am responsible for cloud engineering, DevOps, and on-premises Kubernetes platforms. A big part of my team’s function is looking at leveraging cloud technology to improve existing operations, systems, and workflows, or introducing new ways to deliver our operation.
In January 2022, we were already looking to the end of the season, specifically at the Abu Dhabi Young Driver Test. This is where up-and-coming drivers get the opportunity to experience Formula 1 cars, often for the first time. This allows the Formula 1 teams to analyse the driver’s skills and how they perform in their cars. The test happens right after the Abu Dhabi Grand Prix—with a single setup day before running the test.
Historically, we leave our purpose-built circa 150 tonne Event Technical Centre (ETC), which is our mobile trackside facility. The ETC does all trackside acquisition before forwarding that data to our Media & Technology Centre (M&TC) in the UK for data and video processing. Leaving the ETC at the track for the Young Driver Test means we need to set up generators and large amounts of energy are required for power. Extra staff is also required, resulting in more hotel rooms and car rentals.
Although the Young Driver Test is not filmed live, the team is still required to provide data services such as live timing and driver positioning. Therefore, we require parts of our custom systems for acquisition, processing, and distribution.
This is where AWS comes in.
In this blog post, I cover how Formula 1 uses AWS to leverage an almost fully remote operation to fulfill obligations for the Abu Dhabi Young Driver test, achieving a major step in Formula 1’s remote production effort.
At Formula 1, we look to push innovation and engineer new, more efficient ways to carry out our operation. An example of this was our switch to remote operations during the 2020 Covid pandemic, which was built in just 7 weeks. Our pledge to net-zero carbon by 2030 is one of the key drivers in using innovation to achieve sustainability.
The first challenge was to bring about ease-of-use for our Timing operators at the track. They were the ones operating and consuming the system, which meant we needed to provide a smooth user experience. We looked to Amazon Workspaces for this, as they provided the ideal environment to operate these applications. Amazon Workspaces provides native zero-client functionality, as well as smooth streaming protocols such as NICE DCV. This meant that when we set up the system, it was just plug and play. We also scaled up the Workspaces to use their GPU-based instances for rendering our Timing and Driver positioning pages, which we distribute around the track to provide power to process and render in real time.
For the content delivery element, we leveraged Amazon Elastic Compute Cloud (Amazon EC2) instances because of their scalability and the hundreds of instance types offered. This allowed Formula 1 to choose the compute configuration that was ideal to the custom backend workloads.
With the compute designs in place, we had to look at the network topology for this system to work. In Formula 1, latency is a challenge because of the real-time requirements of our systems. Therefore, for this solution, it made sense to leverage AWS Direct Connect as the bridge between our on-premises infrastructure and the cloud, providing us with that low latency, low jitter, dedicated connectivity into AWS.
Now, we connected the track to our M&TC via our wide area network (WAN), which was already in place following the Abu Dhabi Grand Prix. The connectivity was in place; however, Formula 1 is a big advocate of multicast networking for both video and data. This was a challenge, as the system needed to send and receive multicast traffic. Thankfully, AWS Transit Gateway with Multicast supports this, and allowed us to configure the required multicast groups and add the required elastic network interfaces (ENI) from both Amazon Workspaces and Amazon EC2. Once configured, this functionality worked out of the box, and thus the hurdle was behind us.
The fully remote Abu Dhabi Young Drivers test was a success, with two, 8-hour days of live use running entirely out of AWS, paving the way for Formula 1 remote data operations going forward.
- Infrastructure as Code (IaC). Ability to template this system via IaC tools such as Terraform, allowing us to deploy on-demand when any ad hoc requirements surface. That repeatability element allows us to ensure that same result every time, rather than individually configuring a system each time it is needed.
- Realize sustainability. This is a step towards our sustainability movement as we venture into remote production via AWS. This proven solution has opened the doors to thinking about other solutions that can run in AWS, giving us higher sustainability through efficient compute, as well as reduced trackside freight being sent to the circuit.
- On-demand scalability. This allows us to scale workloads to any resource configuration. Going forward, we know that we have enough resources to meet any demand.
- Predictable pricing. The predictability of the pricing meant that we knew how much the system costs on a per event/day/hourly basis. We easily generated cost predictions for management because of the simple pricing model of AWS services such as Workspaces and EC2.
In conclusion, AWS helped Formula 1 make the leap to fully remote production by enabling real-time virtual desktop infrastructure (VDI) environments through AWS Workspaces, backend compute through Amazon EC2, and networking through AWS Direct Connect. This has now paved a way for Formula 1 innovation, providing new ways in which remote production can be achieved by leveraging the cloud, pushing us in the right direction of net-zero carbon and sustainable operations. As we move forward, we are looking to AWS to leverage these benefits and provide more opportunities that are predictable, reliable, and secure.