Eco-Efficient Cloud Manufacturing: A Paradigm for Energy Consumption Optimization

ABSTRACT: The exponential growth of data center operations has raised significant concerns regarding their substantial energy consumption and its environmental impact. Cloud usage is increasing more and more from companies and individuals. The rising use of cloud manufacturing, innovative way of conducting operations to which companies are joining, presents concerns about its environmental impact. In this context, this thesis work aims at developing a methodological framework to boost energy consumption optimization within cloud manufacturing infrastructures. The procedure starts with a comprehensive energy diagnostic, utilizing indirect energy measurement metrics like CPU usage, RAM statistics, network traffic, and more, to pinpoint energy inefficiencies and hotspots within cloud manufacturing systems. Then, the framework reviews the diverse techniques to be used to improve energy consumption by increasing system efficiency. The framework includes an energy inefficiency allocation analysis module endowed with a root cause analysis for problem characterization. Some possible corrective actions are shown to improve cause inefficiencies which are distinct depending on where the cause is allocated in the cloud manufacturing system. The advantage of the proposed framework lies in its dual capability to not only detect but also mitigate inefficiencies by employing computer-science-based techniques. Through a careful review of system logs and a critical examination of energy consumption trends, the thesis shows how effective the framework is, enabling potential long-term improvements in cloud manufacturing. This work demonstrates how management engineering principles and technology advancements can be combined to improve productivity and sustainability in the context of digital manufacturing. A simulated case scenario covering various relevant scenarios is presented to demonstrate the applicability of the proposed framework.