Presented By O’Reilly and Cloudera
Make Data Work
21–22 May 2018: Training
22–24 May 2018: Tutorials & Conference
London, UK

Real-time motorcycle racing optimization

fausto morales (Arundo), Marty Cochrane (Arundo)
10:0010:30 Tuesday, 22 May 2018
Data science and machine learning
Location: Capital Suite 2/3 Level: Non-technical
Average rating: ****.
(4.67, 3 ratings)

In motorcycle racing, there is a clear trade-off between speed and stability. Riders must decide how they will maneuver and accelerate based on weather, competitor behavior, track conditions, tire conditions, and the geometry of the track. But riders get distracted, overwhelmed, and fatigued. Given all these elements, you can hardly expect them to make the best possible decision every time. However, making the wrong decision can lead to an accident or loss of competitive edge.

Fortunately, there is a long history of implementing automated systems to enable humans to make decisions more quickly. Machine learning, in particular, affords the potential for achieving reaction times far shorter than that of a human for tasks for which we can provide a suitably detailed dataset. Fausto Morales and Marty Cochrane explain how they use a custom software-based edge agent and machine learning to automate real-time maneuvering decisions in order to limit tire slip during races, thereby mitigating risk and enhancing competitive advantage. The solution uses off-the-shelf logging hardware and custom in-house software to capture key sensor data and act on model predictions.

The goal in building the optimization system is to support riders by tracking all the relevant variables associated with their riding and make data-driven decisions. Because slip behavior can vary by rider, bike, and track, the first several laps are used to select and train a model. The completed model then runs on embedded equipment on the bike to limit acceleration and speed in real time.

Photo of fausto morales

fausto morales

Arundo

Fausto Morales is a data scientist at Arundo Analytics, where he works on product development and customer projects. Previously, Fausto worked at ExxonMobil on projects that included environmental remediation, product pricing, and water treatment process modeling. He holds a bachelor’s degree in civil engineering from MIT.

Photo of Marty Cochrane

Marty Cochrane

Arundo

Marty Cochrane is the director of solution architecture for EMEA at Arundo. Previously, he led software development at Statkraft, Europe’s biggest renewable energy company. Marty’s background is in software engineering, with a specialization in industrial applications. In his spare time, he competes professionally around the world racing superbikes. Marty has also been developing his own technical platform that is used in top teams around Europe to develop riders’ skills.