Laps to laptops: Formula One sells its big data know-how

Reuters

By Jeremy Wagstaff

SINGAPORE, Oct 25 (Reuters) - Wander around the pits at aFormula One car race and you're as likely to bump into alaptop-wielding scientist or engineer as a mechanic with aspanner.

And the lessons they are drawing from sensors on F1 tracks,cars and drivers are finding their way into a surprising rangeof industries - from drilling oil wells to making toothpaste.

"By chance or whatever we've ended up that F1 is a verystrong metaphor for how the world is developing around a moreindustrialised Internet," says Peter van Manen, managingdirector of McLaren Electronic Systems, part of a group whichmakes F1 cars. "You take information and you measure things, andfrom that you try to adapt how things behave and flow, so youcan make performance better."

Formula One racing, born after World War II, has longembraced rapid innovation. It can take seven years to get anordinary car from the drawing board to the showroom: An F1 carmay take just five months ... and have new components added toit each race weekend.

"There are very few industries which have a similar abilityto evolve at such pace and bring components or products tomarket at such speed," says Gerard Spensley, AT&T's globalaccounts director for F1.

But in recent years safety concerns and fears it wouldbankrupt itself have forced the sport to adopt tighterregulations, reducing speeds and spending. This has shiftedemphasis from hardware upgrades to real-time tweaks inefficiency and tactics to prise an extra millisecond or two fromcar and driver.

To do that, teams capture gigabytes of data from more than100 sensors on each F1 car, transmitting it back to the pit ordirect to their UK headquarters over high-speed cables. Onceengineers have analysed the data they feed advice back to thedriver - often within minutes or even seconds.

At last year's Brazilian Grand Prix, for example, theInfiniti Red Bull Racing team's UK factory used AT&T's highbandwidth link to assess the impact of an early collision ondriver Sebastian Vettel's car in time to send instructions totrackside engineers ahead of the first pit stop. That gave themtime, AT&T's Spensley said, to make alterations that helpedreduce the risk of further damage to the car. Vettel finishedsixth, retaining his title.

It's this pressure-cooker data analytics which some F1companies say offers an edge not only to racing motor cars butto other industries grappling with big data. The appeal lies inhow F1 teams tackle grabbing information on the fly, figuringout what's important and then converting that quickly into astrategy.

COMPLEXITY

Leading the charge is McLaren, a racing car company foundedin 1963 and now half-owned by Bahrain Mumtalakat Holding Co. Itsapplied technologies unit last month set up its Asianheadquarters in Singapore.

McLaren, for example, has used its experience grabbing datafrom fast moving cars to help build a network of antennae,sensors and masts for San Francisco's Bay Area Rapid Transit(BART). The network is used to collect video surveillance of thecarriages, monitor usage and provide passengers with WiFi.

This is not, says McLaren's van Manen, as easy as it sounds.Data bounces off walls, arrives in the wrong order or drops outentirely. "The complexity is in the detail of dealing with boththe high volume of data and the imperfections in the world."

McLaren is now working on similar projects in Europe.

To be sure, F1 is a small industry and McLaren is beingdriven in part by circumstance to seek new markets. In 2012 thegroup made a pre-tax loss of 2.5 million pounds ($4.05 million).The global market for big data is expected to be worth $23.7billion by 2016, most of it driven by big players such as IBM, Cisco Systems Inc and EMC Corp,forecasts Credit Suisse.

BRAND APPEAL

Not all in the industry are convinced there's much mileagein exporting its expertise.

"There is relevance but our analysis of the marketplace isthat we're far better concentrating very strongly on automotiveand other forms of motor sports," says Paul Newsome of WilliamsF1, whose hybrid engine technology is being tested on buses.

And even those who partner with F1 acknowledge that part ofits appeal is the brand. Paul Marriott of SAP, whosereal-time data analytics product HANA is being road-tested byMcLaren, says the company also benefits from the glamorousassociation with F1. A McLaren car sits in the company'sSingapore office.

But the partnerships run deeper. McLaren is working withU.S.-based IO Data Centers LLC to make data centers moreefficient by modelling likely demand.

"Every plane and car is now driven by software, constantlyadjusting every variable. Taking that thinking and applying itto the data center - where every input is constantly monitoredand controlled - was how the partnership was spun with McLaren,"IO's chief innovation officer Kevin Malik said in an interviewat the company's new Singapore data center.

The way McLaren uses predictive analytics to figure out thebest time to call a racing car in for a pit-stop is helpingBritain's airport controllers anticipate which planes shouldland on what runway. The software is currently being deployed atLondon's Heathrow Airport, according to McLaren.

McLaren is also working with an oil and gas company tofigure out the optimal path to drill through a honeycomb ofwells 10 miles underground. The solution, says McLaren's GeoffMcGrath, is a strategy straight out of racing: measure thecondition of the car, mine the models using historical data andthen advise a course of action - which will change all the timeas new events occur.

MANUFACTURING PIT STOPS

No action is too mundane for F1 to capture and analyse.Teams of mechanics practice changing wheels and tyres dozens oftimes a day, each session captured by video monitors and thenstudied for lessons.

This caught the eye of GlaxoSmithKline Plc, whichasked McLaren to help reduce the time spent switching itstoothpaste production lines. By standardizing materials, makingthe machines easier to handle and reviewing teams' performanceafter every batch, they cut the changeovers from 39 minutes to15.

"Changeovers in the manufacturing world are very like pitstops in an F1 team," says GSK's global manufacturing and supplychief, Roger Connor.

That said, the success of F1 teams in marketing theirexpertise to other industries may have less to do with theircutting edge innovation, than with their high profile work.

"It isn't so much the technology and methods that aredriving, but that these cases are so interesting that people arestarting to pay attention to things like sensors in thevehicle," says Craig Stires, a Singapore-based analyst with IDC."It creates that interest that radiates into other industries."

Whatever the reason, F1's innovation is causing ripples.Providing high-speed connectivity to F1, for example, has helpedcompanies like AT&T and Tata Communications theirservices overall.

Normally, setting up a network for a client would take 10weeks, Tata Communications chief marketing officer JulieWoods-Moss. F1 needs to have it done in three days. "So there'sbeen huge process innovations in the company that now we'restarting to see feeding to mainstream customers."

View Comments (0)