lOgo

guido.tech: Bridging HPC with Real-Time Race Decision Making

Overview

Guido is a proof-of-concept system that connects High-Performance Computing (HPC) with real-time race strategy. Our project simulates how HPC-level analysis could be integrated into a Formula 1 race engineer’s workflow to generate adaptive strategy insights during a race.

How It Works

Data Streaming: A Raspberry Pi acts as the racecar computer, sending batches of telemetry and historical data (via FastF1 API) to a simulated HPC cluster.
Data Enrichment: The HPC node processes incoming data and enriches it with new parameters such as aerodynamics, tire degradation, and fuel efficiency.
Strategy Generation: An LLM interprets the enriched dataset to produce high-level strategic outputs (e.g., pit recommendations, pace adjustments).
Action Layer: The system either:
- Communicates insights via ElevenLabs-generated audio (virtual race engineer), or
- Acts autonomously by simulating control variable changes in the car model.

Vision

Our long-term goal is to demonstrate how AI + HPC synergy can bring predictive, data-driven decision making to real-time systems — from race strategy to autonomous vehicles.