https://racepulse.infoset.co/

RacePulse is an intelligent real time racing analytics platform designed to revolutionize drivers performance combining telemetry data processing, advanced analytics, visualization and live AI coach!

Comment

Inspiration

Everyone who’s spent time in a racing sim knows that driver assists are one of the fastest ways to improve and shave seconds off a lap. That’s exactly what we’ve built, an intelligent coach that talks to the driver in real time, using live telemetry and past mistakes to guide every turn. And it does this not for one car, but at scale across all racers simultaneously .

What it does

RacePulse is a comprehensive platform that empowers teams and drivers with actionable insights. It processes real-time telemetry data to provide:

  • Real time driving analysis: Analyze the fastest possible lap based on historical and real-time data.
  • Turn-by-turn recommendations: Offer braking, throttle, and racing line suggestions for each corner.
  • Performance dashboards: Visualize key metrics like sector times, speed, and vehicle dynamics in real-time.
  • AI coach deciding for live instructions: AI engineer coach to recomend and talk to the driver!

How we built it

RacePulse is built on a robust and scalable architecture:

  • Backend: Python services handle real-time telemetry ingestion using AWS Kinesis and Flink. Machine learning models analyze data for optimal lap predictions and turn recommendations.
  • Frontend: A Svelte-based dashboard, powered by TypeScript on AWS Cloudfront, provides an intuitive interface for visualizing telemetry data and insights.
  • Infrastructure: Terraform provisions cloud resources, ensuring scalability and reliability. Docker containers streamline deployment across environments.
  • Data Pipeline: Telemetry data flows seamlessly from vehicles to the cloud, where it is processed and visualized in real-time.

Challenges we ran into

Building RacePulse presented several challenges:

  • Real-time processing: Achieving low-latency data ingestion and processing required extensive tuning of AWS Kinesis and Flink. We succeeded to have millisecond latency on recommendations.
  • Data accuracy: Filtering noisy GPS data and ensuring precise racing line predictions was a complex task.
  • Scalability: Designing a system capable of handling high data throughput while remaining cost-effective was a constant balancing act.

Accomplishments that we're proud of

We’re incredibly proud of what we’ve achieved with RacePulse:

  • Successfully implemented a real-time telemetry pipeline capable of processing high-throughput data.
  • Developed machine learning models that provide actionable insights for drivers and teams in real-time.
  • Created an intuitive and visually appealing dashboard that simplifies complex data.
  • Built a scalable and reliable infrastructure using modern cloud technologies.

What we learned

RacePulse was a tremendous learning experience. We gained valuable insights on how to translate raw telemetry data into meaningful insights for racing teams. Key takeaways include:

  • Real-time data processing and the challenges of low-latency systems.
  • The importance of user-centric design in creating intuitive dashboards.
  • Applying machine learning to real-world problems in motorsports.
  • Best practices for managing cloud infrastructure with Terraform and Docker.

What's next for Infoset

The journey doesn’t stop here. We’re only at the beginning of what we imagine for RacePulse.

RacePulse is the foundation for a new generation of motorsports analytics, opening the way for major advancements in the way drivers and teams understand performance.

  • Validation in real racing environments: Continuously refining the system through feedback from drivers and teams in actual racing conditions.
  • Enhanced AI models: Bringing in more sophisticated machine learning techniques to deliver deeper, more accurate predictions.
  • Broader integrations*: Expanding compatibility with more telemetry devices and popular racing simulators.
  • Team and Community features: Enable teams to share and compare data, fostering collaboration and competition.
  • Global scalability: Optimize the platform for use in international racing series and events.

We’re genuinely excited about what’s coming next, and we’ll be thrilled to invite you to try the platform live at a racing event soon!!

Built With

Share this project:

Updates