Methodological Challenges of Creating a Next-Generation Machine Learning-Based Game Engine for Generating Maps and Vehicle Behavior

Abstract

The article presents conceptual work in the area of technological challenges for the construction of a next-generation racing game engine. The team creating the game decided to implement the project based on the Unreal solution. The main innovation of the product is to be: realism of driving experience and the use of machine learning elements to improve work under the engine. In terms of realism, the team will make real measurements using race vehicle telemetry devices. The Machine Learning component is intended for two purposes as a map detailing element and for setting the parameters of physics of objects in the game. The paper aims to present the ideal project idea to the critical analysis and thus improve the concept.