Notes – Detecting Safety Car Laps from F1 Laptime Data, Part 1

Looking at some track/lap maps for F1 races, there are several visual cues for when a safety car has been deployed.

For example, here’s a chart from the 2018 Chinese Grand Prix showing when a safety car arrived. The bunching of times at the left show that a full safety car is in track, bunching up the cars. The time-distance that race leader BOT is behind the car at the back in terms of track position increases to show the slower than race pace lap time.

Here’s how the race from Australia 2018 looked. In this case, there was a virtual safety car to begin with – the field doesn’t close up but the time-distance between the race leader and the track position backmarker increases to reflect the reduced laptime:

Singapore, 2017, was a wet race with lots of safety car action. To begin with, the race started under the safety car, and came out again on lap 11. The steady pattern of race trails show there was no safety car around lap 29/30, although there were pitstops, but there was a safety car around lap 40. I’m not sure if a VSC went out first, or whether the teams took the opportunity to pit.

 

So how can we detect full and virtual safety cars? Certainly there seems to be graphical evidence of safety cars – both full and virtual – in the above charts. But can we also identify and detect some numerical signatures, and if so, what form might they take? That will be in the next post…