Farming exergame using webcam skeletal ML tracking and Godot

Abstract: The modern office environment requires employees to perform stationary work. This can cause them several health issues due to being sedentary for long periods of time. An exergame was developed with the aim to remedy these issues by encouraging short motion controlled game sessions during breaks. The game was a farming game developed using the Godot game engine and the ML based skeletal tracking software Mediapipe. The game was played for 86 two-minute sessions where data was collected to evaluate how to use farming game tropes within the design for more movement to yield better results. It was also used to evaluate how the game mechanics could be distributed throughout the game session to promote sustained exertion during an average session. It was concluded that for farming exergames, if the goal is to reward players who exert themselves more, the in-game rewards should not make the game easier to play effectively as they generally do in typical farming games. Furthermore, it was also concluded that interactions happening in a sporadic order help with keeping the average exertion consistent, in addition to the aspect that players always have something to do during the allotted game session time.