Reliability is a key requirement in the aerospace industry. Therefore, structural health monitoring (SHM) applications using strain-field estimation and pattern recognition techniques are in development as an alternative to improve reliability and reduce maintenance costs, promising to ease damage detection on aerospace structures. However, some challenges need to be resolved before real implementation of these techniques. One of these challenges is to uncouple operational conditions changes in strain field from those related directly to damage. In this paper, the development of a strain measurement remote acquisition and transmission system on an unmanned aerial vehicle (UAV) using Fiber Grating Sensors (FBGs) is presented. Before flight testing, ground tests are carried out to emulate the dynamic loads that will be presented during flight. The strain data acquired are processed using unsupervised learning algorithms based on Self-Organizing Maps (SOM) and DS2L-SOM (density-based techniques) in order to classify different operational conditions. The results showed the capability of the system for classifying different load conditions for a UAV's main beam.