HumanIoT – Internet das Coisas centrada no Humano

We are experiencing unprecedented technological advancements, allowing us to extend our tools and appliances with intelligence, sensing and communication capabilities. This idea began with the “Internet of Things” vision. However, the continued advances in the miniaturization of computational facilitates enabled us to go beyond the simple tagging and identification, towards integrating computational resources directly into these objects, thus making our tools effectively “intelligent”. Yet, there is scarce scientific work that considers humans as an integral part of these IoT-powered cyber-physical systems. This proposal intends to discover how we can take advantage of intelligent systems and computational platforms such as smartphones and IoT devices, intensively used by people throughout their daily activity, to build not only an Internet of Things, but rather an “Internet of All”: a socio-technological Internet that considers human parameters like actions, behaviors, skills, satisfaction, emotions, and drives of the ordinary user, the Human, as part of larger scale systems. This project intends to combine the research carried out in the area of Edge AI [1] [2] with research carried out at INESCC and UC in the area of Human in the Loop and corresponding prototypes [A1-A10] to offer an innovative model of Edge AI that integrates the Human in the new Internet. The participation of IPM-UC allows the evaluation of Human psychometric variables. The model proposed within the scope of this project will be based on TL and TinyML, which offer personalized models for inferring but also for training. The HumanIoT model will be evaluated in two case studies, taking advantage of ongoing projects at INESCC. Both case studies are oriented to HiL models, combining physical and emotional user data, and where our HumanIoT solution will allow to take advantage of distributed AI without compromising data privacy. The first case study is extremely important from an academic and social point of view. We have seen great evolutions in the university model, but very little has changed in the role of the "classroom". Consequently, it is crucial to find new pedagogical models, use adaptive learning techniques, so that what is taught by the professor meets the profile of each student, his/her learning abilities and his/her motivations. The second case study has economic relevance and market potential. This case study aims to take advantage of and evaluate the HumanIoT model in optimizing industrial processes in some partner companies like Conduril (Portugal), Intermolde (Portugal) and Vale (Brazil). The latter is a global mining company that stands at the forefront of the natural resources industry. Vale is a leading producer of iron ore, nickel, copper, and various other metals essential for industrial processes worldwide. The objective of HumanIoT is to find new methods of productive efficiency by combining different characteristics of the equipment in the field with the profile and physical-emotional state of each worker.