Virgin Racing's approach to data mixing powers lessons for the future of business

Think of motor racing and what comes to mind is the circus that surrounds Formula One – with its petrol fumes, its noise and accompanying bling. But it’s not the only game in town and one of the newest versions of motor racing may well turn out to be the most interesting. The all-electric Formula E take on the sport has provided a new set of challenges for team bosses, engineers and IT vendors alike.

What sets Formula E apart is not just the fact that it relies on electric motors, but that the set-up allows members of the public to not only support their favourite drivers but give them additional power. It’s the equivalent of a football team gaining a free kick thanks to the strength of its fans’ singing or an America's Got Talent or X-Factor act to win out, not on votes but on the volume of Twitter activity. Perhaps this makes electric car racing the first genuine 21st century sport – which brings with it, 21st century challenges.

All this means that the Formula E racing teams have a particular challenge: first of all they have to make their engines run as efficiently as possible – all the cars start with exactly the same amount of electricity from the same supply, so it’s handled makes a big difference. And having done that, combine this with the support from the fan-base to derive optimal performance. All told, it’s a big technical challenge.

DS Virgin Racing is one of the teams jostling for position in this new sport and is still getting to terms with the intricacies of the sport, which involve making the most of the resources available. As Alex Tai, team principal of Virgin Racing points out:

Formula One teams travel with large numbers of specialists, each with their particular roles to play. There are only 15 people in the Formula E team – and only four of those are engineers.

Electric dreams

The chosen IT partner for Virgin Racing is HP Enterprise, which has both the hardware and software capabilities to deal with the demands of the support. This set-up includes HP Vertica to handle the structured data and Autonomy for the unstructured input (those social media feeds), all of which is being crunched by an array of energy-efficient and speedy Moonshot servers.

Nor can the teams rely on ‘secret sauce’ to work with, no hidden ingredient as they crunch the numbers, syas Tai, pointing out:

There’s one big difference between Formula One and Formula E - all our data is public.

What this does mean is that the IT resources behind each time have to handle considerable amounts of data. Tai is coy about revealing the exact size of the datasets that he has to deal with, but it’s certainly in the order of terabytes.

There are two racks of equipment keeping the Virgin Racing team in business. One that stays trackside to crunch the data during every race and one that stays in ‘mission control’ at Donnington Park. Both of the set-ups use the Moonshot 1500 chassis. The trackside rack has 45 Proliant M170 cartridges while the one based in Donnington Park relies on 30 M170s.

The compact size of the Moonshot range means that it’s well-suited for this flexibility – transporting some of the behemoths of the past wouldn’t be such a trivial task. But this range of servers offers a level of scalability for the type of big data number crunching needed to get the optimum results. The racing team needs results and needs them fast, notes Tai:

Every tenth of a second there’s a different interaction being fed back to the team.

Engineers have to process all that input and crunch that data in minutes, making for an enormous technical challenge, he adds:

We need to the data to get the optimum power/weight ratio.  We looked at a number of IT vendors and HPE stood out because of their analytic capabilities, the depth of their support – their engineers travel with us - and their truly international reach.

While electric car racing seems like a niche venture at the moment, Tai points to some considerable interest in the sport, with many cities clamouring to be included on the Formula E circuit.

This means lugging the servers around the globe in order to have that processing power on hand. It may seem an obvious example of how cloud could be brought into play but Tai says it wouldn’t be practical:

We looked at cloud and it’s something we’d like to do – it was actually in our original spec - but there’s not the connectivity available in cities that we go to.

It’s a facet of cloud that the technology is only as good as the supporting network and Virgin Racing is one of many companies that has experienced that limitation.

My take

Tai suggests that there are pointers for everyday motoring in all of this:

One day, everyone in the heart of the cities will be driving electric vehicles: petrol-powered cars will be treated in the same way that smoking is now.

But that’s not the only example of the principles behind Formula E are going to impinge on everyday life. The Virgin Racing team’s handling of structured and unstructured data is a pointer of the way companies are going to be doing business in future. Not many businesses are going to analysing volumes of data in the short periods of time that the racing teams are, but the mix of database and social media traffic is going to create a few headaches for companies in the years to come.

It’s not going to be easy. The Virgin Racing set-up of two racks of Moonshot servers and the combination of Autonomy and Vertica is not going to be the cheapest option. Given the resources that the race team has to juggle with, it’s certainly the best option but, more generally, it sets out how companies in the future are going to proceed when it comes to mixing structured and unstructured data: it’s a problem that stretches well beyond the world of racing.