
Every tactical option. Lap by lap.
Optimum plan
Lap 30 of 72
Our car, P6: pit lap 44 for softs. Projected P4.
Attack · Undercut
Lap 30| Option | Pit lap | Tyre | Rank |
|---|---|---|---|
| No action | No stop | – | 6 |
| Undercut | 35 | S | 5 |
Race simulation
The optimum plan for every driver, re-run each lap.
Decision analysis
Every tactical option, weighed against that plan and the rivals around you.
Pre-Race
Race, tyres and pace, racing and tactics, set up step by step. Race day starts ready.
Pre-Race
Configuration status
Every input shows where it came from: confirmed by your team, or estimated.
Race
Plan
Every driver's optimum plan, re-run each lap.
Tactics
Attack, defend and respond windows, shown before they open.
Options
Select a window to see each option and where it finishes.
Tactics
Laps 31 to 50
Tactic lanes for our car, laps 31 to 50. Attack: undercut windows open on laps 35 and 44, with overcut and follow watched from lap 35 to 44. Defend: idle. Respond: safety car and VSC responses ready from lap 31 to 35.
AUndercut
AOvercut
AFollow
DPrevent undercut
DCover undercut
DCover overcut
RSafety car
RVSC
Attack · Undercut · lap 35
Gap 3.0 s to the P4 car
No action
P6
Undercut
P5
Respond
A safety car or VSC can close a window in a lap. The response is already evaluated.
Respond · VSC
Lap 55
VSC running since lap 55
Projected finish · baseline P4
Post-Race
Review the reasoning alongside the result, and carry it into the next race.
Decision review
Lap 55
Known at the time
VSC running. Pit now: projected P2. Stay out: projected P3.
Action taken
Pitted on lap 55 for softsAligned
What happened next
Finished P4
Platform and SDK
Run the SwitchPad app as it is, or build your own tactics system on the API and Python SDK.
App
API
Python SDK
from switchpad import Client c = Client.from_env()v = c.replay.get("Zandvoort", 2026, 30, car) v.plan # the optimum planfor t in v.tactics("attack"): a = v.analysis(t) print(a.title, a.output["answer"])Series
Five modelling groups, configured for each championship.
Tap a group to see its levers. Scroll across for every series.
Getting started
30 minutes on a race replay, live or as a link.
Validate it on your past races, with your data.
Connect your live feeds and run it on race day.
Why SIG ML
SIG Machine Learning builds engineering software, powered by machine learning, used by energy operators including Santos, BHP and Origin Energy. SwitchPad is our dedicated platform for race strategy. We build the software and support your team to validate and run it.