Lotus F1 to make faster on-track decisions with cloud
19 June 2014 | 0
Formula One racing team Lotus hopes to virtualise all of its infrastructure to enable it to use datacentres near the 20 racetracks it visits each season instead of shipping hardware track to track.
This would allow the British racing team to make huge cost savings, as well as speed up on-track decisions, according to Thomas Mayer, Lotus F1 Team’s COO.
“We spend millions every year shipping hardware around the world,” Mayer told the Cloud World Forum conference in London this week.
“We want our complete infrastructure virtualised so when we go to a race we move the complete infrastructure to a datacentre that is close to the track.”
Mayer said that virtualising infrastructure would also enable the organisation to carry out more real-time predictions, utilising their data in almost real-time to “make the car go faster” when they need to make decisions during a race.
“We want a two-second window ahead of forecasting, which we are already testing. The technology exists; it is nothing revolutionary, we just to need to use the state of the art technology.”
However, latency has been an obstacle to virtualisation for Lotus. “We want to get it [the lag] to two seconds to make decisions, [but] latency is an issue. If we can [resolve] the latency and security issues, we will do it [virtualise servers]. The technology exists and we will use it to give us a competitive edge and cut costs.”
Security and governance are an added concern Lotus has had to consider while migrating to the cloud and when piloting temporary datacentres utilisation, Mayer said.
“We have highly sensitive data – you could easily gain the 10-year experience level of performance advantage we have so we do not want to lose it.”
To counteract this, Lotus currently encrypts its hosted data to 128 bit encryption – an Advanced Encryption Standard (AES) that the military uses. Lotus’ partners segregate their data pool into different disks so it is fragmented and unusable when separated.
Cloud at 320kph
Last year, Lotus overhauled its legacy data centre infrastructure, setting up a private cloud to increase availability of critical systems as it targets victory at future world championship events.
Many of Lotus’ Formula 1 applications have successfully been hosted in the cloud to allow geographical flexibility.
Mayer told the Cloud World Forum: “I can analyse everything like I did before but I can move my applications first to Malaysia, finish the race, then move to the US — I can travel the world. With 19 or 20 races each season, I need a completely mobile and flexible solution.”
The team currently uses the Microsoft Dynamics ERP tool, which Mayer said will move into the cloud by 2017. It uses Microsoft Azure to host its web pages, and incorporates mobility across these platforms for all their users. It is currently piloting HPC clusters to deal with the high seasonal load and looking at improving its user-facing tools including a consumer application.
Lotus has a total of 550 members of staff who design, build and operate its cars including those for Formula 1. Supporting its wider business are 44 IT staff, including 20 developers focusing on car analytics and configuration.
Margi Murphy, Computerworld UK