Traditional self-adaptive systems research has focused on external contextual aspects such as performance, system reaction to environment. In this paper, we introduce the idea of measuring emotions in order to empower the adaptability of software services at runtime. We present two type of monitoring mechanisms and an adaptive adaptation strategy, which were implemented as part of the HAPPYNESS middleware. A preliminary test using data from the Empatica repository was carried out with the purpose of assessing the goodness of the controller (i.e. inference engine), a component of our middleware. The obtained results were consistent with the expected values. Moreover, we test also the connectivity and synchronization between E4-Wristband and an adaptive mobile application that were used by two volunteer users.