Energy savings with building’s automation is related to the functionality of the building’s automation system, equipment automation level, control algorithms and user-selected operating parameters. It is difficult and time-consuming process to customize each of these algorithms to a level, which would ensure the rational use of energy resources. To increase the adjusting quality that controls building’s engineering system it is necessary to implement new advanced methods, technologies and control algorithms within building’s automation system. This work presents the method of creating self-learning system, which is capable of detecting negative influences of different control algorithms. This self-learning system also provides error detection in engineering systems of the building. The system was designed in Matlab and Simulink programming environments. The article also describes the architecture of designing a complex wireless network system with real time (online) connection, including automatics and building’s engineering systems. This complex system is also capable of collecting data to perform an effective intellectual analysis.