An Open Sensing and Acting Platform for Context-Aware Affective Support in Ambient Intelligent Educational Settings

Supporting learners affectively while carrying out stressful educational activities is an open research issue. It requires the appropriate infrastructure for recognizing emotional states and reacting accordingly in runtime. In this paper, we describe the open platform that we have implemented (named AICARP v2) to detect changes in physiological signals that can be associated with stressful situations, and when this happens, it recommends the learner to relax by delivering modulated sensorial support in terms of light, sound, or vibration at a relaxation breath rate. In this way, by taking advantage of ambient intelligence, the learner can perceive the recommended action without interrupting the learning activity (in this case, practicing the oral exam of a second language). The signal acquisition of the system (which combines sensors from Libellium e-Health platform with others integrated ad hoc) has been compared with a commercial system (J&J Engineering I-330-C2), obtaining similar outcomes as to identifying significant changes in the physiological signals when the learner experiments an emotional reaction. However, the cost of AICARP v2 is much lower, and at the same time, it is open hardware and flexible and, thus, has the advantage of providing runtime data processing. User studies have served to evaluate participants' perception of the sensorial support, as well as to calibrate the delivery rule and to evaluate the effectiveness of the support provided to them.