Williams Martini Racing IT director Graeme Hackland: 'If we're not at the forefront of technology, we won't win'

Hackland tells Danny Palmer how data analytics and a partnership with BT have helped transform the F1 team and how machine learning, IoT and cloud bring further benefits

During the team's 1990s heyday, legendary drivers like Alain Prost, Nigel Mansell and Damon Hill won Formula One World Championships behind the wheel of a Williams Racing car.

But by 2013 those days seemed long ago, as the team struggled to score just five points in one of the worst seasons in their history.

That led to team owner Sir Frank Williams implementing a full-scale transformation of his operation, which by the time the 2014 season rolled around was rebranded Williams Martini Racing.

As part of that transformation, Graeme Hackland was appointed as the team's IT director in January 2014. With 18 years of experience in Formula 1, including over three years as IT director at Lotus F1, Hackland was hired to help move the team up the grid.

"I was asked to come in and do IT transformation, look at technology across the company and how we can use technology better to make the car quicker, to take on teams better funded than we are," he tells Computing.

The transformation appears to be working as Williams went on to finish third in the 2014 Constructors' Championship and looks set to repeat the achievement this year.

Hackland is supported by an IT team of 20 and together they are responsible for "supporting every part of the lifecycle of a Formula 1 car". That ranges from the design and manufacture of the car to supporting drivers Valtteri Bottas and Felipe Massa and the trackside team on race weekends.

One area Hackland immediately singled out for improvement was how Williams collected and used data.

"Data was being generated by the car but we weren't getting it to the engineer in the time they needed to make decisions," he explains. The answer was a project he describes as a "digitisation of race team strategy" that was designed to help better harness data on tyres, fuel and other vital areas.

Hackland also identified networking and bandwidth infrastructure as areas that needed to be improved.

"We weren't able to move data around the world from the track back to the factory as quickly as we wanted to, so we moved onto our partnership with BT in order to get that data in real time, that makes a big, big improvement for the team," he explains.

BT boost

Why did Williams opt for BT? "You've really got to go with one of the big telcos," argues Hackland, adding that the partnership with BT "just felt right".

"Their CEO and directors met all our senior executives and there just seemed to be a connection between what BT are trying to do, innovate to stay ahead of their competitors, and what we're doing," he says.

One of the first jobs BT tackled was to speed up network performance so that engineers at the track could faster access to the data they need.

"Latency was a big issue, especially in Asia, Australia, it was a big problem for us in terms of video and voice, and BT have helped us to solve that," says Hackland.

"Compared to the start of the year, the difference in performance is like chalk and cheese; our engineers are able to work differently now," he says.

The upgrade has made it easier for engineers at the track and back at the team's Oxfordshire factory to collaborate effectively, essential for a team with a relatively small headcount.

"The more people we can have working with the data, the more brains you can add to the problem of finding a way to get round the track quicker," says Hackland.

Rivals such as Mercedes and Ferrari might have three times the number of staff Williams do, but Hackland believes that by leveraging technology his team is able to match their rivals. One technology Williams is looking to exploit more is cloud computing.

"We do some cloud but I want to get us to the stage where we have a hybrid model," Hackland tells Computing, before describing how deploying extra computing power as and when it's needed can bring advantages.

"When the car is on the track, you need to have all of the data in systems locally at the track, but when the car's not on the track, we should be able to just switch seamlessly into a cloud model that will give us more computing capability," he says.

For Hackland, the ultimate aim is to be "computing constantly". However, it isn't realistic to install the equipment at trackside, but it can be drawn from the cloud.

"We can use cloud capability to do that extra analysis and then when that engineer wants to look at the results, they're there already, that's where we're heading with cloud," he says.

"You get more capacity than we could put in our data centre, it's more reliable than what we could build at our own site," Hackland adds.

An internet-connected Formula 1 car?

That demand for more computing capability is only going to increase as Williams Martini Racing looks to exploit the power of new data-generating technologies.

Connected devices and the Internet of Things (IoT) could, for example, play a major role in the design and manufacture of new cars. Rival F1 team Infiniti Red Bull racing are also examining how connected devices can provide an advantage.

"I think where the Internet of Things is going to play a big part is putting sensors in all sorts of areas: the factory, manufacturing, the design side of things. I think we can see improvements to traditional CAD drawings using sensors and technology in manufacturing," says Hackland.

He also believes machine-learning algorithms could also play a crucial role.

"A Formula 1 car is in constant evolution. Although we only make four or five cars a year, after every single race that car changes. There are times when you'll manufacture something while the designer is redesigning it," he says.

It's here that machine learning and artificial intelligence could play a role by causing the machine that is manufacturing the part or the car to automatically react if there's a change made to an associated design file.

"We think with machine learning, AI and the Internet of Things, we might be able to join up that lifecycle so if a designer opens a part and starts to amend it, the manufacturing capability will know that's the next job," Hackland says.

"If a designer is redesigning a part the likelihood is you're not going to use the original part, and the machines can start to make decisions about what they're manufacturing," he adds.

And that might just be scraping the surface of how the Formula 1 team will eventually harness the power of machine learning.

"In the F1 lifecycle of the car, we think machine learning is going to benefit how we manufacture the car," says Hackland.

"The decision-making that's made on a daily basis around what you put into the machine shop, the optimum way of designing next year's car, is fairly manual now. We're automating, but machine learning will help us with how we manufacture the car.

"There are lots of areas we can apply machine learning to that will feed into Formula 1 and make us more competitive," he adds.

When it comes to the technology companies providing the software, Hackland is less interested in the size or pedigree of the vendor than what they can offer.

"I'm trying to make sure we don't only talk to the established companies that everyone knows but also start-ups," he says, describing how the team's partnership with BT helps in this area.

"Working with people like BT, who go out to Silicon Valley and talk to the start-ups, they can give us access to technology we'd have difficulty finding," Hackland says, adding: "I'm keen not to just use established technology."

Indeed, Williams is examining the possibility of heads-up displays for the pit crew in order to make pit stops quicker.

"I've been talking to a prospective partner about how we could instrument the pit crew in order to help them. Our CTO wants to see pit stops under two seconds and to get the human below two seconds is going to need a loT and technology will have a big part to play," says Hackland.

Ultimately, if using innovative new technology can help Williams win races, the team wants to use it.

"If it's the reliability of our car, we're not going to take any chances. But why not try bleeding-edge technology in other parts like manufacturing and design and get a jump on the other teams before they use it," says Hackland.

"If we're not at the forefront of technology as a team, we won't win. If we don't keep up, we won't beat the bigger teams with more money, more people, we've got to take them on by being smarter."