CIO Interview: Graeme Hackland, Williams Martini Racing
Williams Martini Racing CIO Graeme Hackland discusses network infrastructure, data analytics and the future for wearable tech and augmented reality in Formula 1 motor racing
It is the Thursday before the 2016 British Grand Prix, and as a summer breeze blows around the pit lane at Silverstone, Williams Martini Racing CIO Graeme Hackland looks on as the official Formula 1 scrutineers cluster around the disassembled shell of Valtteri Bottas’ Williams Mercedes FW38, his arms folded, looking for all the world like the lord of his domain.
And well he might. Two and a half years ago, when Hackland joined Williams from rivals Lotus – the world of Formula 1 being the sort of tight-knit place where those with the right skills can spend decade-spanning careers flitting from one team to another – IT was not, by his own admission, a priority for the family-run team.
Over the course of 16 years at various incarnations of Lotus, Hackland had essentially built the team’s IT infrastructure from the ground up. But when he arrived at Williams, he faced a major new challenge.
“Formula 1 has always been geared up around rapid iteration and a process of trial and error. The car is never finished, and what we weren’t doing at Williams was doing the same thing with our IT,” he reflects back in the team’s motorhome.
“Before I joined, I think IT at Williams was put in a corner, told to keep the lights on and do it cheaper this year than last year. But I think there’s now an understanding that technology and IT can make a car quicker and add to performance.”
Williams has won only one race in the past decade, but with an epic history and multiple drivers – including Alain Prost, Nigel Mansell and Damon Hill – winning Formula 1 championships in its cars, performance has clearly been an issue of late. Could this now be on the turn?
Formula 1 fans will already know that a move to Mercedes-powered engines in 2014 has paid off in terms of speed and reliability, with drivers Valtteri Bottas and Felipe Massa now able to challenge for podium places.
New networking deal
Behind the scenes, Hackland has also been pedalling frantically. Last year he signed a major new networking deal with BT to use its high-performance networking services for secure, superfast communication and collaboration, and cloud-based fixed and mobile voice.
He has also deployed a 100Mbps multi-protocol label switching (MPLS) network that delivers symmetric speeds between Williams’ headquarters at Grove near Oxford, and 21 Formula 1 racetracks all over the world.
One immediate impact of this network refresh is a reduction in the amount of IT equipment the team hauls around when on location. Before last year, it took four racks of converged server, storage and networking hardware with it, but now it only has to take two. This mobile, mini-datacentre sits in the back of the cramped garages at Silverstone, surrounded by racks of carbon-fibre components and piles of wet weather tyres.
I want to stream all our data back to the cloud to allow everyone to access it from there, whether at the factory or at the track
Graeme Hackland, Williams Martini Racing
But for Hackland, even this small saving is not good enough. “In three years, I don’t want it to be there at all,” he says. “I want us to be running exclusively in the cloud. I want to stream all our data back to the cloud to allow everyone to access it from there, whether at the factory or at the track.”
Williams is already encrypting all its data in the Microsoft Azure cloud, and has engaged business services specialist Avanade – itself a joint venture between Microsoft and business consultancy Accenture – to ensure elements of the team’s IT can run at the track on race weekends or in the cloud the rest of the time.
Moving to the cloud will also enable Hackland to eliminate an extra expense. Whereas at Grove he can run equipment for five to seven years in a standard equipment lifecycle, at the track things are always being moved around and getting bumped, and so the hardware needs a full upgrade every two years or so.
Partners for profit and strategy
The increased reliability of the underlying infrastructure means that lately, Hackland has been exploring new strategies around how Williams uses software applications to gain a competitive edge on the track, which is where supplier Avanade comes in.
For a sponsor, partnering with a Formula 1 team such as Williams is more than just a means of blowing some advertising budget and an excuse for some corporate hospitality every now and then, although it is partly both of those things. Perhaps more importantly, it should be seen a learning opportunity for both parties.
For example, Unilever, which sponsors Williams through its Sure and Rexona brands of anti-perspirant, helps out by applying its production-line knowhow to the team’s small factory, while in return it learns more about the importance of streamlining decision-making processes in a business.
Elsewhere, Williams’ expertise in aerodynamic design is obviously of use to aerospace specialists, while its knowledge of energy-efficient technologies has brought it new clients in renewable energy and electric vehicles. Here, it has even lent its experience to Formula E, the world’s first fully electric-powered motorsport series.
Besides a decal on the car, Avanade also gets to apply the team’s stores of knowledge to its own business and, in return, it has taken a front seat when it comes to helping Hackland make best use of new applications.
New capabilities
It has already enhanced the car production process at Grove with new capabilities around quality control and reporting that are enabling Williams’ design and manufacturing teams to work more closely together than ever before to weed out components that don’t work.
“We’re expecting that this will mean designs that were rejected in the past won’t now get to that stage, that manufacturing will find it easier to make the components that design is pushing their way, and so there will be less wastage,” says Hackland.
“These non-conformance reports are mostly a process change, and they are built on Microsoft SharePoint as well. We don’t want exotic platforms that are specific to Formula 1 – we want everyone to be able to take what they produce from us and the knowledge and know-how of what we do, and apply it to other customers.”
We think we’ve gone ahead of other teams with some of the predictive analytics
Graeme Hackland, Williams Martini Racing
At the Austrian Grand Prix earlier in 2016, Williams deployed new strategic predictive analytics capabilities, again provided by Avanade. With new rules governing tyre use in force this year, both Bottas, Massa and their race engineers have more choice in which tyre compounds they use, and this has made racing a lot more complex strategically.
For Pat Symonds and his teams, the new analytics capabilities mean he is better able to calculate race strategy on the fly, and even adjust pit strategies if, for example, a certain tyre compound shows heavier graining.
The new strategy system is a source of particular pride for Williams. Hackland reckons it has given them a distinct advantage over the likes of Mercedes, Ferrari and Red Bull.
“We think we’ve gone ahead of other teams with some of the predictive analytics,” he says. “We don’t think they have that capability yet, and that’s the beauty of it. The partnership and using the offshore model gives us access to much more resource than we could have put to it ourselves, which means we can jump ahead of the other teams rather than just following what they’re doing.”
Marginal improvements
Of course, there remains much work to be done to get the drivers to the top step of the podium, and so Hackland is constantly looking ahead to bring in new technologies wherever possible.
One aspect of IT everyone is keen to discuss right now is the internet of things (IoT). However, this is one area that is actually rather old hat in Formula 1: the first IoT-type sensor was fitted to a racing car in 1979, before anybody had heard of the internet of things, or even the actual internet. It was a 64kb data-logger and it measured three aspects of the car’s telemetry. It also took 20 minutes to download one lap’s worth of data.
So with the concept of the IoT well established in the sport, Hackland is now looking to human instrumentation as the next step.
The Renault Sport Formula 1 team explains why it is considering a refresh of its trackside converged infrastructure setup.
Microsoft Dynamics tools drive the Lotus F1 team’s data analytics in ratcheting the performance of its cars, engineers – and business.
“Avanade is running a new project to instrument our pit crews,” he says. “Pat [Symonds] has a vision to get our pit stops down under two seconds on average. We were the fastest team on pit stops for the first nine races of 2016, and our average is two point two. If we want to get below that, we have to get down to the very fine margins of human performance, so IoT and wearables have a real role to play.”
Hackland cites an incident that occurred at the Austrian Grand Prix as a good example of how technology can be used to improve human performance.
Valtteri Bottas accidentally stopped his car slightly short of his box – the painted marks on the ground that guide the driver during a pit stop – which meant that every member of the pit crew had milliseconds to react and adjust their positions.
“I think there’s a role for wearable technology, maybe something that would give an audible signal in the mechanics’ ears, which would be based on the GPS data from where the car is positioned,” he says.
“We can already see [from real-time braking telemetry] that the car is going to stop short, so the mechanics need to move slightly to the right, and they get a beep in their ear to tell them to do that. There’s a lot we can do in terms of human performance. That is where the future gains are going to come from: small, tiny leaps around the margins, saving fractions of a second.”
The virtual future
While the fans can rest assured that Formula 1 won’t go fully autonomous, the world of virtual and augmented reality (AR) is starting to make its presence felt in the paddock as well.
“This is a big area we are looking at right now,” says Hackland. “Take the mechanics working on the cars in the garages. If they want to check a technical diagram, they have to print out a 2D drawing and hold it up to the car and work out how the part fits. We see potential here for AR to help them see where a part should fit, its orientation, and so on.”
Overall car design should also benefit from this. Right now, Williams is preparing to switch more and more of its design staff onto next year’s car – and with big changes to technical regulations on the way for 2017, this is no small task.
“We do a good job of building a 60% wind tunnel model to simulate aerodynamics,” says Hackland, “but we can’t do a lot of visualisation of those models. I think we can make design decisions much quicker if we can visualise it instead.”
