Back to blog
overviewarchitectureai-director

Sim RaceCenter: AI-Powered Broadcast Production for Sim Racing

An overview of Sim RaceCenter — what it does, who it's for, and how it transforms live sim races into professional-quality broadcasts.

·Sim RaceCenter Team·4 min read
On this page

Sim RaceCenter is an AI-powered broadcast production platform for sim racing. It transforms live simulated races into professional-quality broadcasts by combining real-time telemetry data, automated camera control, and intelligent production decisions.

What It Does

Sim RaceCenter automates the work of a broadcast director for sim racing events. During a live race, it continuously analyzes telemetry data from the racing simulator (currently iRacing) and makes real-time decisions about camera angles, scene transitions, and broadcast narrative -- the same decisions a human TV director would make during a motorsport broadcast.

The platform consists of three main components:

  1. Race Control -- a web-based command center where Broadcast Directors configure race sessions, manage driver profiles, set up OBS scenes, and monitor live broadcasts
  2. Director -- a desktop application that runs alongside the racing simulator, executing camera commands, managing OBS scene transitions, and polling for AI-generated broadcast sequences
  3. Race Control API -- a serverless backend that houses the AI Director pipeline, manages all configuration data, and serves as the bridge between the web interface and the desktop application

Who It's For

Sim RaceCenter is designed for Broadcast Directors -- people who produce and direct live sim racing broadcasts. These are typically:

  • League administrators who want professional broadcast quality for their racing series
  • Content creators streaming sim racing on platforms like YouTube and Twitch
  • Esports organizations producing competitive sim racing events
  • Sim racing enthusiasts who want to create TV-quality race coverage

The Broadcast Director does not need to be a technical user. A Broadcast Agent (AI assistant) guides them through session setup, driver configuration, and broadcast management using conversational flows.

Core Capabilities

AI-Powered Camera Direction

The AI Director pipeline uses a two-tier architecture:

  • Planner (runs once at session start): Analyzes the session configuration, available camera capabilities, and connected integrations to generate a library of parameterized sequence templates. Each template describes a type of broadcast shot (e.g., "Battle Close-Up", "Leader Spotlight", "Field Sweep") with variable placeholders that get filled at runtime.

  • Executor (runs continuously during the race): On every poll cycle (every 15-90 seconds), analyzes live telemetry data -- race positions, gaps between cars, flag states, incidents -- and selects the most appropriate template. It fills in the variables (which driver to focus on, which camera angle to use, how long to hold the shot) and returns a concrete sequence of broadcast commands.

Real-Time Telemetry Analysis

The platform ingests telemetry from the racing simulator via an Event Hub pipeline:

  • Car positions and lap times updated every few seconds
  • Gap analysis between cars to detect battles (cars within 1 second)
  • Flag state monitoring (green, yellow, red, white, checkered)
  • Incident detection for instant replay opportunities
  • Pit stop tracking for strategy-based camera decisions

OBS Integration

The Director application controls OBS Studio (Open Broadcaster Software) to:

  • Switch between pre-configured scenes (race view, battle cam, pit lane, standings overlay)
  • Trigger scene transitions with appropriate timing
  • Manage overlays (lower thirds, leaderboard, timing tower)
  • Handle replay sequences for incidents

Multi-Driver Support

Sim RaceCenter supports both single-driver and multi-driver broadcast sessions:

  • Solo sessions: Focus on one driver's perspective, switching between cockpit, chase, and external cameras
  • Multi-driver sessions: Cover the entire field, rotating between drivers based on race narrative (battles, leaders, incidents)
  • Driver rotation fairness ensures balanced coverage across all participants

Architecture

Sim RaceCenter runs on Azure cloud infrastructure:

  • Azure Static Web Apps hosts the Race Control web interface (Next.js with server-side rendering)
  • Azure Functions (Flex Consumption) powers the Race Control API
  • Azure Cosmos DB (NoSQL, serverless) stores all configuration and runtime data
  • Azure Event Hubs handles real-time telemetry ingestion
  • Application Insights provides observability across the entire stack
  • Google Cloud Vertex AI (Gemini models) powers the AI Director pipeline
  • Microsoft Entra ID handles authentication; managed identity secures all service-to-service communication

All infrastructure is provisioned and managed through Terraform (Infrastructure as Code).

Brand Identity

Sim RaceCenter's visual identity evokes a professional motorsport control room:

  • Apex Orange (#FF5F1F) -- primary action color, highlights, live status indicators
  • Telemetry Blue (#0EA5E9) -- data visualization, AI-related elements, information displays
  • Race Control dark theme -- deep black backgrounds with elevated grey surfaces for a cinematic, high-contrast aesthetic
  • Inter typeface -- clean and modern for UI elements; monospace for telemetry data

The tagline is "Orchestrate the Chaos" -- reflecting how the platform brings order to the dynamic, unpredictable nature of live racing.