Augmentative and alternative communication (AAC) encompasses a wide range of methods that replace and/or complement speech of individuals with complex communication needs. Predominant AAC methods rely on the interpretation of purposeful gestures; however, such methods limit the solutions in terms of versatility and portability. Moreover, the cost of predominant high-tech AAC systems is generally very high. An alternative AAC solution, based on encoded modulated breathing, is being researched to address the shortfall in this area. The system undergoing development has been validated with the participation of over 39 non-speech disabled participants in two experimental protocols to test modalities of picking up breathing patterns, such as microphones and pressure sensors. The current results show a mean systematic reliability of 93% by utilising machine learning and dynamic programming to learn and recognize the breathing patterns. The results demonstrate that an improved breath-activated AAC solution could be sought in the future.