How Pitchside AI Works

What Pitchside AI Does

Pitchside AI is built to turn football recordings into useful match output. Instead of only storing a video file, it analyses the footage and generates stats, highlights and player moments.

Step 1: Record the Match

Teams can record football matches using one phone or two phones. The best setup is usually above head height near the halfway line. Two phones can point toward each goal, while one phone can work if it captures enough of the pitch.

Step 2: Computer Vision Reads the Footage

Pitchside uses computer vision to understand football actions from the video. The model is trained on small-sided football rather than professional broadcast footage, because grassroots games have different angles, lighting, movement and visual cues.

Step 3: Pitchside Detects Events and Stats

The first release is designed to detect goals, assists, saves, passes and tackles. Pitchside can assign those stats to teams and individual players, which makes leaderboards and player comparisons possible.

Step 4: Highlights Are Generated

Pitchside is also being built to produce full match highlights. This matters because most grassroots players do not want to rewatch a whole recording. They want the best goals, saves, assists and moments.

What Is Still Improving?

Pitchside is still in beta. Accuracy is improving, and the current upload process can take longer than ideal. As more footage is processed, the learning algorithm should get better across different lighting, pitch types and recording angles.