The amount and diversity of data generated by the sensors available on current Wearable Computing Systems (WCS) impose the adoption of advanced, efficient (and often customized) data analysis methods, including Machine Learning (ML) / Artificial Intelligence (AI) and multi-sensor information fusion. With the diffusion of Cloud Computing and ARM/MCU architectures in the last decade, commercial WCS have gradually been introduced into the market, resulting in vertical solutions with poor pro-grammability, privacy, and smartness. To make WCS truly smart and Cloud/smartphone independent, we propose COCOWEARS: a framework for COntinuum Computing WEARable Systems. COCOWEARS relies on (i) sound theoretical foundations with a hardware/software system co-design methodology based on simulations, (ii) tailored AI/ML algorithms, and (iii) development tools to enable the engineering of next-generation WCS. A use case about the monitoring of operators' activities within the smart factory is provided to exemplify our approach, which is, to our knowledge, the first solution fully exploiting AI technologies at different levels of granularity, complexity, and user proximity, able to push towards massive adoption of WCS in the immediate future.
COCOWEARS: A framework for COntinuum COmputing WEARable Systems
Aloi G.;Frustaci F.;Gravina R.;Guerrieri A.;Savaglio C.
2025-01-01
Abstract
The amount and diversity of data generated by the sensors available on current Wearable Computing Systems (WCS) impose the adoption of advanced, efficient (and often customized) data analysis methods, including Machine Learning (ML) / Artificial Intelligence (AI) and multi-sensor information fusion. With the diffusion of Cloud Computing and ARM/MCU architectures in the last decade, commercial WCS have gradually been introduced into the market, resulting in vertical solutions with poor pro-grammability, privacy, and smartness. To make WCS truly smart and Cloud/smartphone independent, we propose COCOWEARS: a framework for COntinuum Computing WEARable Systems. COCOWEARS relies on (i) sound theoretical foundations with a hardware/software system co-design methodology based on simulations, (ii) tailored AI/ML algorithms, and (iii) development tools to enable the engineering of next-generation WCS. A use case about the monitoring of operators' activities within the smart factory is provided to exemplify our approach, which is, to our knowledge, the first solution fully exploiting AI technologies at different levels of granularity, complexity, and user proximity, able to push towards massive adoption of WCS in the immediate future.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
