Compute Continuum in Bioengineering: Enhanced Motion Capture and Real-Time Data Processing on Cloud-Edge Infrastructures

The constant evolution of cloud-edge computing, correlated with advancements in bioengineering technologies, has opened new opportunities for real-time data processing and analysis in critical applications. This dissertation explores the integration of motion capture (MoCap) systems with Internet of Things (IoT) architectures, leveraging cloud-edge platforms such as OpenStack and Stack4Things to address scalability and efficiency challenges across domains like healthcare and smart cities. At the core of this research is the development of MocapMe, a framework designed to optimize motion capture and analysis through markerless technologies powered by deep learning models. This framework can allow a wide range of applications, including clinical rehabilitation, sports performance analysis, and real-time animation, showing notable accuracy and system responsiveness improvements. Moreover, the MocapMe integration within a Compute Continuum architecture aims to minimize latency, promoting real-time feedback for critical motion capture tasks in medical and sports contexts. Enabling technologies such as LoRaWAN and peer-to-peer (P2P) networking are investigated to provide robust communication in distributed systems, even in resource-constrained environments. Moreover, the dissertation explores the dynamic virtualization and management of IoT resources through the I/Ocloud paradigm, ensuring seamless scalability and efficient data processing across cloud and edge environments. The research also delves into the Compute Continuum prospect of integrating IoT, cloud, and edge computing in diverse applications. Case studies cover a range of use cases, from smart city infrastructures to healthcare scenarios, including sit-to-stand (STS) analysis, demonstrating the practical advantages of the developed technologies in these domains.