A successful multi-agent system requires the intelligent agents to perform within a dynamically complex environment where proper and quick response in a cooperative manner is a primary key to successfully complete a task. This paper proposes a non-deterministic decision making method using electric fields and high-level decision making. Different layers are designed, defined, and implemented for the software architecture with focus on system adaptability, sustainability, and optimization. Consequently, a software architecture is proposed in this paper to complement the AI algorithms. The proposed architecture aims to provide a well-structured and managed system for control, behavior, and decision making of multi-agent systems. The proposed decision making approach in this paper is based on layered artificial intelligence implemented using vectorbased fuzzy electric fields and a decision tree. Furthermore, an approach to model the world which, in this paper, is called Agent Relative Polar Localization is introduced. This world model is based on fuzzy measurements and polar coordinates. In order to optimize the overall performance of the system learning methods have been introduced to the system. The proposed system in this paper has been implemented on soccer robots to evaluate the performance of the system. The results show that the proposed system implemented on the soccer robots is reliable and robust.