Fusion Q: AI-Powered Pool Training Software

DesignCoursego watch the original →

Fusion Q is a projection-based pool training system that uses computer vision to track ball movement, provide real-time shot analysis, and enable remote multiplayer drills.

System Architecture and Capabilities

Fusion Q is a software platform designed for pool training that integrates computer vision with overhead projection to overlay digital feedback directly onto a physical pool table. The system requires a PC, a camera, and a projector. The software uses a custom ball-detection model to track individual balls, calculate trajectories, and provide visual indicators for shot paths, including neutral, stun, draw, and multi-rail angles. Users can also apply real-time shaders to the table surface for visual feedback during drills.

Training and Community Features

Beyond basic tracking, the software includes a drill editor that allows users to create custom training scenarios in 3D or by capturing the current table layout. The platform records every session, enabling users to review shot-by-shot replays, analyze statistics, and share their performance data with a community feed. A standout feature is the "Instant Recall" mode, which saves snapshots of ball layouts after every shot, allowing players to reset the table to a previous state to practice specific positions repeatedly. The system also supports internet-based multiplayer, where users can challenge others to drills and games in real-time.

Hardware and Accessibility

The developer emphasizes affordability, noting that the system can run on a $400 mini-PC or a used gaming PC with a dedicated GPU. The setup requires a standard camera, such as the Logitech MX Brio, and a projector, both of which can often be sourced used for approximately $150 to $200. The project is currently in a pre-release phase with a Discord community providing support for hardware configuration and environment setup.

  • #demo
  • #ai
  • #dev-tooling

summary by google/gemini-3.1-flash-lite. probably wrong about something. check the source.