All Formula One teams compete for the same goal and are pitted against the same competitors but the playing field is far from level. In 2016, the four biggest teams had budgets over twice that of Williams Martini Racing. Despite this, Williams generally maintains performance on a par with its richer peers.
Information management and information technology are now key areas in which Williams has been able to level the financial inequality. The structure of the Williams Group itself helps in this process. This includes both F1 and the Williams Advanced Engineering division which works with partners and clients from many sectors where F1 specialist knowledge can be applied.
This side of the business adds an extra dimension to the capabilities of the F1 team. While it is important to protect intellectual property in F1, Graeme Hackland, IT director at Williams, says: "F1 data has a relatively short life, a couple of seasons and then no one else is really interested. But for Williams Advanced Engineering, a lot of the technology and intellectual property they are working with is very valuable so we needed to prove to our customers that we are protecting their information and we know where it is. The audit capability we have with our technical partner Dtex means we can tell what’s been printed, what’s been emailed, what’s been stored on any devices and whether someone is on our network or at the factory or if they are mobile. It’s important to our audit capability and to some of our key customers who want us to do those regular inspections."
Last year Williams Advanced Engineering signed a £17m contract to design and manufacture a system for distributing power and data around the British Army’s next generation armoured fighting vehicle. Projects like this require a high level of security. The technology to achieve this is implemented group-wide so Williams’ Formula One division benefits as well. "It’s also key to F1," Graeme says, "at a time like now when next year’s rules are so different the teams are going down different routes it’s really important that we all protect our information. In collaboration with our partner Dtex, we spend a fair amount of time looking at where our data is and getting alerts if there’s any misuse of data and dealing with it on the day it happens."
These IP fears are not unfounded after a 2007 "spygate" case involving McLaren and Ferrari. F1 is a small world and there is a movement between teams – Graeme came to Williams from Lotus in 2014 – and he says: "I’ve got to prove to the technical team that their data is safe and that if someone has left the team nothing has gone with them."
IM on the track
Formula One cars can have 150-300 sensors collecting data. A number of secondary systems analyse this data or combine it with other data sources. For example, an app developed by one of Williams’ tech partners, Avanade, receives data from the trackside data warehouse integrating sensor, weather and other data feeds in real-time to isolate the impact of tyre status on performance.
Graeme said that the management of this car data had defined the sport for decades: "I think teams put huge effort into this area, ever since cars first became connected in the late 1970s." He says Williams "makes sure that each engineer has a specific specialisation at the track and they don’t need to see all the data. Only two performance engineers watch everything, but pretty much everyone else filters it down to what’s relevant for them".
Much of this is done with alerting tools that allow engineers to set baselines and watch specific parameters. He says "a lot of that is about making sure that the right piece of data gets to the right engineer at the right time".
These, however, are becoming increasingly complex, hence increasingly involved technology partnerships. "We’ve deployed a new strategy system this year which is light years ahead of what we have been using at Williams before. You can usually do a pretty good job of your own cars’ data because you’ve got all of it. But it is incredibly complex especially when you start to add in competitor analysis."
There are pressures between providing live data and data for future research and development. As a result, the car is set up differently to accommodate different data gathering processes ahead of a race. Graeme says: "On the Friday of a live weekend we capture a lot more data from a car than we do on a Saturday and a Sunday. That data feeds into the rest of the factory: so to the wind tunnel, to the vehicle science, the vehicle technology guys and into our drivers’ simulator. We use the Friday to capture R&D data and then Saturday is all about getting the car getting ready for qualifying for the race. Then you go racing on the Sunday."
Similarly, reference to historic data is kept to a minimum on the day. Graeme says: "On the day itself we’ve got people back at the factory who will have done some work with historical data. That is in the lead-up to a race. A big part of a race engineer’s job is to be doing that analysis across multiple years in the weeks before a race. Often by the time they’ve got to the track they’ve got a pretty good idea of how they want to set the car up based on everything we’ve learned over the previous years. Then it’s all about learning from the data that’s been generated now."
Technology and compromise
Graeme says in F1: "There’s lots of stuff that I’d love to play with and see whether it has an application but we’re resource constrained and time constrained so we’re more likely to come at it with technology that is already established."
However, he adds: "If we just use established technology only then we’ll effectively be following the other F1 teams and some of them are bigger than us and have more people and it won’t give us a competitive advantage."
The group structure enables Williams to compete with better funded rivals at the cutting edge of technology: "In F1, we come very much from the position 'we have a problem how can we solve it' while Williams Advanced Engineering are looking for technologies that they can exploit – taking knowledge and know-how from F1 and applying it to all sorts of other interesting industries – as such they will be more proactive."
Some of this advantage comes from the larger partnerships with the likes of BT, which links the UK factory to the track-side race team, and Avanade, who "drive how we move data around and how we improve workflows and how we improve our strategy systems at the track."
But start-ups are also key players in the information arms race with other teams. Again these companies are usually accessed via Williams Advanced Engineering and Graeme says: "If we can get access to that kind of technology before its mainstream and before the other F1 teams are aware of it, we can get a competitive advantage. Those are the ones we don’t put in the public domain… because right now the other teams aren’t aware of them. We’ll talk about the work we’re doing with BT and Avanade, but we don’t disclose the things we’re doing with other niche companies."
While some advances remain under the radar, other mainstream information juggernauts are approaching. For example, Graeme sets 2020 as potential date for the arrival of effective 3D printers on the trackside. This capability is likely to disrupt existing information management systems because the timetable for working on the car changes dramatically. Graeme says: "In the past, when we’ve experimented with 3D printing, the parts just weren’t strong enough to put on an F1 car."
But he added: "I’ve seen the first 3D printed bicycle in carbon so they’re getting to the point…where you can apply it in a motor racing environment. I think by 2020 we should be looking to have 3D printing manufacturing capabilities at the track so the designers can work up until the Thursday night before the race instead of a week or two-weeks before as we currently do."
Other areas of interest in the mainstream like machine learning and artificial intelligence are filtering their way into F1.
Graeme says: "Where you have repetitive data we want to try and automate that especially on some of the real-time decision making that we’ve got at the track for F1. Some areas are obvious candidates, like the rulebook. When there’s an accident into the first corner they are likely to red flag it and I think a role that AI has in the future is to look at that. We currently rely on a couple of humans to do that but we think there’s a future for putting that into AI so it can learn from the decisions that the FIA has made in the past and predict what decisions they might make in the future."
Graeme believes there is a lot of potential in this for F1 and the wider group: "Some of the areas, even for automation robotics around your help desk – when people call in looking for triage on help desk calls – a lot of it is repetitive so could we use a robot to filter those calls. If we could prove it in IT, we could expand it to faults that happen at the track with the car. We could have a robot answering that call and then at the point where you need human intervention it will learn from what the human did and the next time someone calls it’ll refine what it asks and keep refining that process. I think that will apply to an F1 context at the track with a car and car faults and parts that we are shipping out to the car."