Formulae One

Adeline Shin and I created a website (formulae.one) that aims to present Formula 1 data in interesting and creative ways. We use data from specific races, season wide metrics, and various other types to create insights into the fascinating world of Formula 1.

Races are analyzed through the lap times, gaps relative to certain drivers throughout the race, and comparisons between drivers. Historical qualifying data is used to find linear/quadratic fits that can tell us about the performance trends of certain drivers or teams over a season or multiple seasons. We also feed this data into machine learning models in order to make predictions for upcoming races.

A novel section of the website revolves around lap simulation of F1 cars (formulae.one/lapsimulation). This is based on my research work at the University of Michigan (see more). The user can pick a track, and parameters such as the maximum engine power and lift/drag coefficients, and see how they change the performance of a hot lap around the track. This technique is based on optimal control theory applied to a trajectory optimization problem. Various vehicles with different performance parameters are simulated, followed by fitting this data to a radial basis function based surrogate model. This is what allows the user to slide design parameters back and forth.