Análisis en Tiempo Real para el Internet Industrial de las Cosas

The rapid expansion of the Industrial Internet of Things (IIoT) has led to the largescale deployment of interconnected sensors, machines, and cyber-physical systems capable of generating continuous streams of heterogeneous data. While this transformation enables intelligent automation, predictive maintenance, and data-driven optimization, traditional centralized cloud-based architectures are unable to satisfy the strict latency, scalability, and contextual-awareness requirements of modern Industry 4.0 environments. Addressing these challenges requires distributed, adaptive, and real-time data processing mechanisms capable of operating across multiple computational layers. This doctoral research proposes a context- and situational-aware collaborative architecture for real-time heterogeneous data processing in IIoT ecosystems. The primary contribution is the design and implementation of Atmosphere, a unified edge–fog–cloud framework that integrates multi-agent systems, Complex Event Processing (CEP), and Event-Driven Service-Oriented Architecture (ED-SOA). In the edge layer, autonomous software agents perform local filtering, rule-based reasoning, and immediate decision-making, reducing communication overhead and latency. The fog layer incorporates MQTT-based messaging and CEP engines to aggregate, correlate, and analyze distributed event streams in near real-time, enabling coordinated responses among heterogeneous nodes. The cloud layer provides large-scale analytics, long-term knowledge extraction, global orchestration, and strategic optimization. The architecture ensures bidirectional communication, interoperability, and dynamic adaptation based on contextual information. The second contribution validates the proposed architecture through the design and development of an IoT-based, solar-powered camouflaged robotic system inspired by the Fennec desert fox. The robotic platform operates as an intelligent edge node capable of autonomous data acquisition, context-aware decision-making, and energy-efficient operation in harsh and isolated environments. Experimental evaluations demonstrate improved responsiveness, scalability, and operational efficiency compared to traditional centralized monitoring approaches. Overall, this research advances the state of the art in real-time IIoT analytics by delivering a scalable, distributed, and context-aware framework that bridges heterogeneous data sources, intelligent coordination mechanisms, and practical deployment. The proposed solutions contribute to the realization of resilient, adaptive, and intelligent industrial ecosystems aligned with the principles of Industry 4.0.

Author dc:creator

• Zouai, Meftah

Advisors dc:contributor.advisor

• Ortiz Bellot, Guadalupe
• Kazar, Okba

Statement dc:rights

• Attribution-NonCommercial-NoDerivatives 4.0 Internacional

Licence dc:rights.uri

• http://creativecommons.org/licenses/by-nc-nd/4.0/

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/10498/39800

OAI identifier oai:identifier

oai:rodin.uca.es:10498/39800