Machine Learning With McLaren

[MeganBowers]

Alteryx and McLaren Racing have a strong partnership as advanced analytics is driving a new era of racing.

To go along with our latest Alter Everything podcast episode where we hear from Trevor Schulze, Alteryx’s Chief Information Officer, and Zak Brown, McLaren’s CEO, we wanted to share a demo of how Alteryx’s technology can be applied to racing data.

Check out the video below, where Vishal Soni (@vsoni) walks us through an Alteryx Machine Learning use case: determining which car configuration will result in the fastest lap time by optimizing the aerodynamics of a Formula One car.

Vishal demonstrates data setup, model setup, getting insights from the model results, simulation, exporting the model, and automation in the cloud.

To learn more about Alteryx Machine Learning, check out these articles:

The Power of Automated Machine Learning

Machine Learning for the Data Analyst

Why Alteryx Machine Learning

To learn more about McLaren + Alteryx, here are some resources:

5 Brilliant Ways McLaren Racing Uses Data to Meet Their Goals (That You Can Use, Too)

Tune Up Your Analytics with McLaren Racing

Formula 1 and Business: 4 Key Similarities

If you want to try Alteryx Machine Learning to answer your own optimization questions, you can start a free Alteryx Analytics Cloud trial here.

[CailinS]

I assume this podcast / demo will be on repeat for you  @DanM 🏎

[VishalChothani]

hi

[MxS]

Great podcast and demo -- especially interesting to see this use case with such a high profile sport!

Perhaps someone from McLaren, or any other industry, might weigh-in with a comment about a critical challenge with this type of use case: how do you use Alteryx to optimize the racecar for multiple criteria, not just minimizing lap time? For example, I might want to minimize lap time, and simultaneously minimize lap time variance (assuming more than one lap). Ideally the racecar should run low lap times, consistently!

Formulating a goal function comprising of these two dependent variables (average lap time, lap time variance) is a possibility, but this gets complicated quickly with the addition of each dependent variable. Furthermore, a real optimization use case has to account for variation in the independent variables -- no point in identifying a point solution, only to discover any minor deviation in the input values causes the outputs to go off a cliff.

We've solved these real-world challenging problems, but unfortunately have do it outside of Alteryx. I am curious about the following:

1. How does a race team (or any other industry) formulate a probabilistic optimization for competing multiple criteria?

2. Is it possible to implement the solution in Alteryx?

[MeganBowers]

@Ken_Black do you have any thoughts on the above comment from @MxS? I thought of you since you are solving tough racing data problems with Alteryx :) 

[Abdulrazzakblog]

As someone deeply interested in the intersection of AI and sports, I was excited to see Intel's AI platform making waves in athlete identification and development. This technology showcases the power of data-driven insights in predicting and enhancing athletic performance. In a recent article I wrote, I delve into how Intel’s AI could be the key to discovering the next Olympic champions. You can read more about the AI’s potential impact here: https://trendingcoolnews.com/intel-ai-technology-could-identify-the-next-olympic-superstar/ . I'm curious about what other applications you all see for AI in sports!"