Early damage detection remains one of the priorities of the structural health monitoring systems in the task of continuous monitoring. In this kind of systems different approaches can be used, however data-driven systems are requested because the information from the sensors is obtained directly from the structure in real operational and environmental conditions. Some of these approaches makes use of acousto-ultrasonics (AU) techniques, which offer the possibility of inspecting large areas of structures, by using a piezoelectric active sensor network. However, these kind of inspection systems are affected by the variations in the environmental conditions. In this sense, is a need to still working in more a nd better da ma ge detection techniques. This pa per descr ibes a hea lth monitor ing methodology combining the advantages of guided ultrasonic waves together with artificial immune systems as a pattern recognition technique to determine the effects of the temperature in the damage detection process, in addition, a sensor data fusion with the data from different temperatures is proposed as a hefty baseline to consider the healthy structure under different temperature conditions and discarding the resultant false positives by the changes in temperature. Experimental results are included to demonstrate the temperature effects and how the methodology improves the damage detection capabilities.