For the last year of university, we were required to conduct a research project over the course of the year. With an interest in level design, procedural generation and really wanting to make a roadtrip game, I took on the ambitious task of producing a simple driving game with procedurally generated environments. In terms of research, I was interested to work out if a system could be produced to adapt procedurally generated environments to player preferences that an integrated system could identify. 

​

Over the course of the project, I conducted research to identify various approaches to procedural generation and how adaptation and personalisation could be used to create better game experiences for players. Afterwards, with a lot of trial and error, I produced a simple on-the-rails endless driving game in Unreal Engine 4. Through the use of UE4's Blueprints, I was able to produce a straightforward road and environment generator which would populate the sides of the road with environment objects based on 1 of 6 predefined environmental zones. The player would then drive down roads and be presented with new roads and environments and be given the option to accept or decline the new road. As the player continues to drive, making decisions along the way, a preference system would progressively identify and score the environmental zones based on the player's decisions and duration. Based on the scoring, the system would then increase the likeliness of the main predicted preferred zone.

​

A summarised version of the research, project and results can be found in the poster below.