Pit Rho is a real-time, in-race strategy tool used in the NASCAR Cup Series. Pit Rho ingests competitors' laptime and pitstop data into a series of machine and deep learning models that predict a wide range of race variables, including competitors' behaviors, and recommends optimum race strategy on every lap for any car in the field based on those predicted behaviors. This robust cloud-based application performs more than one million calculations per lap while maintaining near 100% reliability under varying data pipeline conditions.
Ingest, clean, and store over 100,000 data points on every lap from telemetry, laptime, and pitstop streams.
Machine and deep learning models that adapt every lap to changing track conditions; the models project cars' future laptimes, positions, and outcomes.
Optimum and predicted strategy and chassis adjustment recommendations for any car on any lap, updated in real-time.
Asynchronous architecture to ensure real-time delivery of strategy. Rapid release cycles guarantee that the latest and greatest technological, motorsports, and mathematical innovations are included.
Extra-trees classifier and Bayesian regression machine learning algorithms create 30,000 decision tree nodes per racecar pairs.
Replay, review, and analyze previous races and create any combination of race scenarios in Pit Rho's Simulation Tool.
Adapt a modern technology stack for seamless integration of telemetry data into Pit Rho.