Designing Pervasive Game Elements to Handle Uncertainty using Fuzzy Logic Systems

• Vitor A. C. C. Almeida UFPI
• José R. M. Viana UFPI
• Ricardo de A. L. Rabêlo UFPI

Resumo

Pervasive games utilize contextual data from the player as a fundamental game input. Processing this data, acquired by sensors subject to noises, imprecisions and errors, is an important task in order to guarantee a consistent gameplay experience and prevent player frustration. Thus, this paper presents an approach to pervasive game development capable of handling the uncertainty of sensor data. The proposed approach consists of applying fuzzy set theory to transform the game input into fuzzy sets, by mapping the range of sensor readings into descriptive linguistic variables (i.e., qualitative values) and their respective membership functions. Then, the linguistic variables are utilized in the definition of fuzzy rules – if...then rule statements – describing the behavior of one or more game elements. Finally, in order to output a value which describes the game element’s behavior, we employ a fuzzy inference system based on the fuzzy rules definitions. To illustrate our approach, we present a pervasive game prototype that uses location and movement contextual data to control the pace of the game. By employing fuzzy systems in the development of pervasive games, we expect to abstract much of the complexity of handling imprecise and uncertain sensor data during the implementation of the game logic and facilitate the design of rule-based behavior by allowing game designers to compose rules using linguistic terms instead of numerical values.