Find, Fuse, Fight: Genetic Algorithms to Provide Engaging Content for Multiplayer Augmented Reality Games

In Augmented Reality (AR) mobile games, several technical aspects are still partially under-explored, thus limiting the creativity of game designers and the spectrum of possible uses of AR. As a result, too often AR is used only to superimpose in a static way predefined digital content to real scenarios. In the present work, we have started to tackle this issue by designing a game to overcome the limited interactivity among players and the somewhat static use of AR on resource-limited devices (i.e., cell phones). In particular, we have designed and prototyped FFF: Find, Fuse, Fight, a game that supports multiplayer mode, offers a more creative use of AR, and demonstrates that Procedural Content Generation (PCG) techniques could be effectively exploited for introducing a higher degree of variability both in the content and in the gameplay, even on devices far less performing than a standard PC. In particular, we developed a prototype that exploits Genetic Algorithms (GAs) to create new content and apply meshes deformation to 3D models in real-time. We have used such content to prototype a mobile game that features AR battles among creatures in an online multiplayer environment. The prototypes have undergone a performance test to evaluate the feasibility of AR multiplayer games with generated content, collecting encouraging preliminary outcomes.