Digital Twin · AI · Energy

GIANT

Framework based on Digital Twin and Artificial Intelligence for the Dynamic Management of Datacenter Resources.

The project

Data Centers (DC) are the beating heart of the digital era, essential for meeting the growing demand for advanced digital services. The GIANT project aims to develop a framework based on Digital Twin (DT) and Artificial Intelligence (AI) to optimise the management of datacenter resources, reducing energy consumption and implementing predictive maintenance strategies.

The GIANT platform will integrate AI, IoT, and physical models to build an interactive digital twin that reflects the real-time operational state of the datacenter. Through a continuous flow of information between the physical and virtual worlds, the Digital Twin will be able to predict anomalies, optimise resource allocation, and minimise energy consumption.

This project aims to innovate current datacenter management, replacing a traditionally reactive approach with predictive management, enabling cost reduction and more sustainable use of energy resources.

Technologies used

Artificial Intelligence and Machine Learning

For forecasting energy consumption and detecting operational anomalies.

IoT (Internet of Things)

For real-time monitoring of the datacenter's operational conditions.

Digital Twin

A faithful digital representation of the datacenter, capable of simulating operational scenarios and optimising resource configuration.

Objectives and expected impact

The GIANT project aims to improve the energy efficiency of datacenters, reducing the carbon footprint and increasing the sustainability of the ICT sector. Specific objectives include:

Energy optimisation

Targeting a Power Usage Effectiveness (PUE) below 1.4.

Predictive maintenance

Identify and prevent anomalies and failures in critical components, reducing downtime and operational costs.

Waste reduction

Resource management system based on predictive data and advanced analysis.

The expected impact is significant not only from an operational efficiency perspective, but also for improving environmental conditions and reducing emissions, in line with the sustainability objectives set by the European Union.

Open-source dataset

As part of the GIANT project, a targeted field data acquisition campaign was carried out to collect detailed information on energy consumption and datacenter workloads. The data is made available by Key4 Srl as open-source, downloadable in CSV format with technical documentation.

The dataset can be used to train machine learning models for data analysis, predictions, and anomaly detection.

Download

Partnership and collaborations

The GIANT project is the result of collaboration between Key4 Srl and WPS Srl, two companies with established experience in technological innovation and engineering consultancy. Key4 Srl handles the design and implementation of advanced Industry 4.0 solutions, while WPS Srl contributes its know-how in advanced software solutions for system monitoring and automation.

Project funded under the SISSA cascade call as Spoke 9 for the Mezzogiorno area, within the research programme of the i-NEST – Interconnected Nord-Est Innovation Ecosystem.

Project code: 2D1C2C7ED7