Tech Research

Hosting della sala

Daniele Pucci
Daniele Pucci
13 jun 12:50 - 13:30
40 min
Renewable energy sources, as vital as they are volatile, are at the center of new distributed energy production systems, such as Renewable Energy Communities (RECs). However, the limited controllability of renewables can lead to power grid instability problems, creating disruptions for the country's civil and manufacturing sectors. We will explore a Reinforcement Learning solution designed to optimize energy flows in RECs, maximizing environmental, social, and economic benefits for both communities and the entire country.
13 jun 14:20 - 15:00
40 min
Artificial Intelligence poses significant challenges to professionals and organizations. Some of these are technical challenges, related to the search for ever greater efficiency in model training and speeding up, for which the use of supercomputing is a key factor. Others concern the imagination of new ways of interacting with Artificial Intelligence: we will explore in particular anticipation and creativity as facilitators of innovative experiences.
13 jun 17:00 - 17:40
40 min
We live in an era where AI is extremely expensive and energy-intensive: large language models (LLMs) have opened up incredible opportunities, but at what cost? Is it really the only way to rely on the use of omniscient LLMs so that AI can assist us in everyday life? We want to discuss how Embedded AI is sometimes a more effective and sustainable alternative.
13 jun 18:00 - 18:40
40 min
In this presentation, I will describe how the vast amount of data on our daily behaviors (social interactions, movements, purchases) and the development of innovative models in the fields of Artificial Intelligence (such as deep learning and generative AI models) and complex systems are enabling the emergence of a new science dedicated to the study and design of cities and urban environments. For example, studies will be presented demonstrating how machine learning algorithms and new data sources can be used to estimate the socio-economic conditions of a neighborhood, predict crime levels in an urban area, and understand which features would allow the design of vibrant places capable of stimulating social interactions and creativity. These findings open the doors to a new way of studying cities and human societies through computational tools and with a depth and scale never before possible.