Data, data everywhere
Want to know how to go faster? Easy – just look at the data. But how, and which bit? Bob Gray tears himself away from his telly to explain…
You’re a team manager with some cash and want to cut lap times, but do you buy more horsepower, better suspension or a more experienced rider? ‘Without data you may as well throw darts at a dartboard – you’ve no idea what will help you more,’ says Bob Gray, the technical director of DAO Racing who run Dean Harrison in BSB and on the roads. ‘However, data doesn’t tell you what you should do – or what you need – it just gives you information to improve your decision.’
Sounds easy enough. ‘But with modern bikes you can record almost everything,’ says Bob, ‘and if you do that, you’ll drown in numbers. Better to identify a problem, then identify what we need to know to solve it, then measure that.’
Got that? Good. You’re now ready to improve your performance using noughts and ones…
Choose your sensors
A basic set-up will have sensors measuring wheel speed, suspension movement, throttle position, brake pressure and gear position. ‘If you put that on an inexperienced rider you can make big gains through big simple chunks of time – for example, just by being in the right gear,’ says Bob. ‘The better and more experienced the rider and team, the less you use those basic channels. You still need them, but you also need other points of reference.
‘So, for example, if they’re backing into a corner, you can’t rely on the wheel speed sensors because the rear is locked (or airborne), and the front is banking into the corner so is running on a smaller circumference. You need some measure of true ground speed. We use GPS and GLONASS – the American and Russian constellations – which are generically called GNSS and typically accurate to 0.1 kph. But that isn’t the real problem.’
Hold on to your hats, this is where it goes all PHD. ‘With engine braking, there is an excess of “negative” power from the engine. We could solve that by opening throttle a lot, but that would make the rider extraordinarily grumpy. We need to work out how much throttle we need in order to achieve the right amount of slip. That means knowing the engine characteristics as well as speed.
‘If you get those from mapping on the dyno and GNSS, the problem then is what wheel speed you actually need because the rolling circumference of the tyre changes all the time depending on load, lean angle, temperature, speed and wear. To some degree this can be modelled by measuring lean angle, suspension position, tyre pressure and temp, but then you still need the speed over ground element to work out how many rpm the wheel “would” be doing. It’s complicated.’ Bob loves an understatement.
Use a tactical lie now and then
Though data is complicated, it’s not as complicated as humans (particularly the racing variety), which means all manner of subtlety is required. ‘Say they come in and report that another rider pulls two bike lengths out on a corner and they need more power,’ says Bob. ‘But when you look at the data you see they’re not opening the throttle very much. So power isn’t the problem – it could be suspension, or we might actually reduce the power, knowing that it will give them the ability to open the throttle more. When they come in and say “that’s much better, it drives so much harder,” the trick is to keep a hold on your tongue.’
Or it could be something else entirely. ‘I’ve had exactly this scenario where we found that because the rider had broken his wrist, he wasn’t physically able to go to full throttle, even though he thought he was.’
And finally, if all else fails, deploy psychological tricks. ‘There’s a thing in racing called the healing shelf. So, for example, you tell a rider that the shock has been changed and serviced, when really it’s just been wiped with a rag and put on the healing shelf. Then you refit it and because the rider thinks it’s different they ride better.’
Beware false feedback
‘When a bike comes in you check the basics – is it overheating, is the fuel pressure ok, is the lambda [ie the air-fuel ratio] where you expect, has the suspension bottomed or topped out? If you see something, you then ask the rider how the bike was going into one corner or whatever – you try not to lead them to what you suspect. It’s also worth double checking by asking the same question in a different way later on. Sometimes riders can be confused, and not all are happy saying “I don’t know” or “I can’t tell”. The worst thing is to get false feedback, and for us to spend hours chasing a problem that isn’t there.’
‘Though data is complicated, it’s not as complicated as humans (particularly the racing variety)’