Augmented intelligence with voice assistance and automated machine learning in Industry 5.0
Alexandros Bousdekis, Mina Foosherian, Mattheos Fikardos, Stefan Wellsandt, Katerina Lepenioti, Enrica Bosani, Gregoris Mentzas, Klaus‐Dieter Thoben
Abstract
Augmented intelligence puts together human and artificial agents to create a socio-technological system, so that they co-evolve by learning and optimizing decisions through intuitive interfaces, such as conversational, voice-enabled interfaces. However, existing research works on voice assistants relies on knowledge management and simulation methods instead of data-driven algorithms. In addition, practical application and evaluation in real-life scenarios are scarce and limited in scope. In this paper, we propose the integration of voice assistance technology with Automated Machine Learning (AutoML) in order to enable the realization of the augmented intelligence paradigm in the context of Industry 5.0. In this way, the user is able to interact with the assistant through Speech-To-Text (STT) and Text-To-Speech (TTS) technologies, and consequently with the Machine Learning (ML) pipelines that are automatically created with AutoML, through voice in order to receive immediate insights while performing their task. The proposed approach was evaluated in a real manufacturing environment. We followed a structured evaluation methodology, and we analyzed the results, which demonstrates the effectiveness of our proposed approach.