Using Speech Data to Automatically Characterize Team Effectiveness to Optimize Power Distribution in Internet-of-Things Applications

This paper focuses on a fresh paradigm where human actions and Machine Learning meet to maximize system performance of Internet-of-Things Edge (IoT-E) with Humans-in-the-Loop applications. To optimally allocate resources, like the available energy, this paper explores the challenges of bridging the semantic gap between team dynamics and team efficiency, so that the more effective teams are given higher priority in resource allocation during operation. The paper proposes methods to interpret team activities using transformer models, like DistilBERT, to process team interactions conducted through speech, and then to utilize the extracted insight to characterize team dynamics. Based on these characteristics, a dynamic power distribution scheme was designed to allocate the available power to teams with higher effectiveness. The results show that the proposed method can improve power allocation in IoT-E applications.