MOTUS - Automated Analysis and Prediction of Human Movement Qualities

The project will develop new methods aimed at achieving radical improvements in technologies for the analysis and prediction of human movement qualities. Particular emphasis is placed on automated detection and tracking of movement qualities fundamental to nonverbal communication between humans. Imagine that we must pass a heavy object to a partner. In this case, anticipatory adjustments to maintain postural equilibrium will involve torso and lower limb adaptations that will dramatically differ if the object is light (i.e. adaptations limited to the arm). Most importantly, our partner will read that information early during the interaction and prepare to expect a heavy or light object. These subtle but fundamental communicative signals are summarized in the concept of Origin of Movement (OoM), i.e. the part of the body perceived by an external observer as the joint in which the movement originates. This project will investigate theoretical developments in the automatic detection and tracking of OoM and related novel applications of machine learning methods and architectures (e.g. hierarchical clustering, matrix completion, and graph neural networks) to human movement analysis during social interactions at different spatio-temporal scales.

P.I.: Giorgio Gnecco

Assegni di Ricerca 2023
Regione Toscana 

ROBOFARM - Development of edge computing tools for robotic platforms in precision agriculture

The project aims to develop edge computing methods optimized for processing data from various types of sensors on board robotic platforms in the context of precision agriculture (project in collaboration with Sigma Engineering).

R.U.: Giorgio Gnecco

PRIN 2022

MAHATMA - Multiscale Analysis of Human and Artificial Trajectories: Models and Applications

The project will develop new mathematical and computational models of expressive trajectories for under-investigated senses (touch) and scales (public-address sound). Supramodal models of trajectories will be derived through cross-sensory translation by means of experimental phenomenology, and validated through controlled experiments. Explainable, feature-based models will be developed as extensions of existing models, and will inform the development of featureless data-driven models. Applications will be developed, as proofs-of-concept for the trajectory models, in the areas of sport and performing arts, with implications in a variety of fields, including culture- and art-enabled motor reactivation and rehabilitation, and navigation help for sensory-impaired people.

U.R.: Giorgio Gnecco

THE_Spoke 3

THE Tuscany Health Ecosystem - Spoke 3 dedicated to Advanced Technologies, Methods and Materials for Human Health and Wellbeing

"THE - Tuscany Health Ecosystem" is one of the 11 innovation ecosystems financed at the national level for the PNRR and is the result of a project proposal submitted by the University of Florence, as lead partner, in line with one of the strategic lines of the Region of Tuscany. THE is the only ecosystem dedicated to Life Sciences, which focuses on the needs of the population. The project has received funding totaling EUR 110 million from the Ministry of University and Research. THE's challenge is to make Tuscany the 'health region'; by pushing research towards applications and companies to grow technologies dedicated to health and wellbeing. The ecosystem will devote part of the funding to provide funds for applied research to companies and spin-offs on the basis of special public calls for tenders, as well as the temporary hiring of about 300 new young researchers.

Spoke 3 is dedicated to Advanced technologies, methods, and materials for human health and well-being.

P.I.: Massimo Riccaboni


Innovative and safe microwave-based imaging technology to make breast cancer screening more accurate, inclusive and female-friendly

Breast cancer (BC) is the most common cancer in women worldwide, affecting 1 in 8 women. Mammography is the gold standard technology for breast screening, which has been demonstrated through different randomized controlled trials to reduce breast cancer mortality. However, it has limitations and potential harms, such as the use of ionizing radiation, breast compression and performance restrictions due to the intrinsic nature of X-rays. In particular, breast density is a restrictive property that can prevent breast cancer detection in mammograms of women with radiographically dense breasts. Other existing techniques (MRI, Ultrasound, biopsy) also suffer from drawbacks. The overall aim of the MammoScreen proposal is to generate evidence about the use of MammoWave (a technology developed by UBT) as screening technique in population-based programs promoted by National or Regional Health Systems, to reach a revolution in breast screening. To do so, the consortium aims to confirm that MammoWave reaches sensitivity >90% and specificity >95% in BC detection on 10000 study participants undergoing regular screening programs. MammoWave uses safe non-invasive and non-ionizing microwave signals, does not apply any compression to the breast and is very effective with dense breasts. A comprehensive health economic assessment will be undertaken in this project and innovative way to implement patient engagement approaches is sought. An effective policy makers’ engagement plan will be carried out to ensure that MammoWave is recommended as a screening approach due to the benefits that it brings to women and healthcare systems. This action is part of the Cancer Mission cluster of projects on ‘Prevention, including Screening.

P.I.: Massimo Riccaboni

Contributi Liberali 2022

Applicazione di tecniche di Matrix completion per la definizione di economic and financial recommender systems

Il progetto mira a velocizzare l'impiego di una specifica tecnica di machine learning supervisionato (matrix completion) per l'analisi della complessità economica di Paesi e città, mediante l'applicazione di opportune tecniche di parallelizzazione.

P.I.: Giorgio Gnecco


Sviluppo di metodi di machine learning per la stima del valore Shapley e di sue generalizzazioni

In questo progetto si intendono investigare metodi di machine learning per la stima del valore Shapley e di sue estensioni, con particolare riferimento a giochi su grafo, modellanti ad esempio reti di trasporto o interazioni tra giocatori all’interno di una squadra.

P.I.: Giorgio Gnecco


Increasing accessibility of online board games to blind and visually impaired people via machine learning

The project aims at surveying accessibility of board games in the case of blind and visually impaired people, and proposing developments based on machine learning and sonification techniques. As a proof of concept, a specific board game (Quantik) is considered.

P.I.: Giorgio Gnecco


Economic Shocks on the Tuscan Economy: 

international competition, energy crisis, and the COVID-19

The project aims to identify and study the phenomena and conditions that can bring out vulnerability and risks of poverty and marginalization of people and territories following recent pandemic and geopolitical events, with particular attention to international competition and rising prices of raw materials and energy. To achieve this goal it combines insights from industrial, labor, energy, and international economics.

P.I.: Francesco Serti


Advanced Robotics and enabling digital Technologies & Systems 4.0

ARTES 4.0 è uno degli 8 Centri di Competenza selezionati dal Ministero dello Sviluppo Economico nell’ambito del Piano Nazione Impresa 4.0 ed è per questo soggetto abilitato alla presentazione di bandi di finanziamento. Grazie ad ARTES 4.0 le imprese possono proporre progetti di innovazione ed essere selezionate per ricevere il contributo da parte del MiSE. Il Competence Centre ARTES 4.0 nasce per associare partner universitari, Enti di ricerca ed istituti di formazione ad elevata qualificazione, fondazioni e aziende partner a carattere innovativo al fine di fornire alle imprese (in particolare alle MPMI) tecnologie e servizi dedicati a rispondere ai loro bisogni mediante attività di orientamento, formazione, progetti di innovazione, ricerca industriale e sviluppo sperimentale.

P.I.: Marco Paggi/Nicola Lattanzi/Massimo Riccaboni

Programma "Galileo" 2021

Automatic movement analysis techniques for applications
in cognitive/motor rehabilitation

The project applies artificial intelligence techniques to medicine by contributing to the design of systems for the automated analysis of full-body human movements for cognitive/motor rehabilitation. The proposed approach has an interdisciplinary nature, being based on motion-capture techniques and advanced computational methods grounded on graph-theoretic and game-theoretic approaches. The project is focused on the automatic analysis of the origin of movement. The research has applications, among others, in the fields of physiotherapy and sports science.

P.I.: Giorgio Gnecco


Trade-off tra numero di esempi e
precisione in varianti del fixed-effects panel data model

In varie applicazioni economiche, ingegneristiche, fisiche, e mediche, si richiede di approssimare una funzione a partire da un insieme finito di esempi supervisionati e rumorosi. Tale problema è studiato, tra gli altri, dal machine learning supervisionato. In alcuni casi, la varianza del rumore sull’etichetta assegnata a ciascun esempio può essere ridotta, aumentando  il costo/tempo di acquisizione di ciascuna supervisione. Questo motiva lo studio del trade-off ottimale tra precisione e numero di esempi. Lo scopo del progetto consiste nell'estendere al fixed-effects generalized-least-squares (FEGLS) model l’analisi di tale trade-off, già studiato in precedenza per modelli più semplici.

P.I.: Giorgio Gnecco

Programma "Galileo" 2019

Both deforestation and pollution are key factors in global climate change. This is because forests are natural pollution sinks able to capture carbon dioxide emissions from the atmosphere and convert them into oxygen, which both humans and animals can breathe safely. Unless proper countermeasures are taken, deforestation represents a big problem for climate: by cutting down forests without replacing the trees that are removed, the absorption efficiency of such carbon sinks is reduced, with possible catastrophic consequences. To counteract these effects, suitable policies are needed, such as reductions in the emission levels, and reforestation.

P.I.: Giorgio Gnecco

PAI 2019:

Prosociality, Cognition and Peer Effects

PRO.CO.P.E. is a multidisciplinary research project funded by the IMT School for Advanced Studies Lucca that studies how the likelihood of prosocial behavior is affected by the mode of cognition, the structure of social interactions and their interplay. PRO.CO.P.E. contributes to the research-based design of policies fostering prosociality.

P.I.: Ennio Bilancini

PAI 2018: VeriOSS

VeriOSS: a Blockchain-based Bug Bounty Platform

VeriOSS is an internal project, funded by IMT School For Advanced Studies under the PAI junior program. Its mission is to imagine a novel solution for a fair, reliable and efficient market where ethical hackers can disclose OSS vulnerabilities. VeriOSS will develop a smart contract-based platform for bug disclosure and reward payment. It offers a fair trade support to boost bug bounty programs and their effectiveness.

P.I.: Gabriele Costa

PAI 2018: Ecopoly

Regional business clusters as economic polymers: mathematical modelling, forecasting, and optimal policy design

The project aims at characterizing the inter-sectoral and intra-sectoral dynamics and business performance of regional economic clusters with the purpose of forecasting possible future evolutions and of suggesting optimal investment policies to public decision makers. In the spirit of a truly multidisciplinary approach, we will exploit computational methodologies drawn from systems engineering (mathematical modeling of discrete systems used in computational materials science to characterize complex polymer chains, machine learning algorithms, numerical optimization, and feedback control design techniques) and methods of strategy and management science (business modeling, balance sheet analysis, competitiveness analysis of companies and clusters). The project will consider the yachting industrial cluster located in Region Tuscany as a specific case study.

P.I.: Nicola Lattanzi

PAI 2018: Future job

Technological change, soft skills and future high skilled jobs

In this project, we put soft and cognitive skills at work to develop a truly interdisciplinary approach to better forecast future occupations and job market dynamics. First, we aim at creating metrics that are theoretically sound and suitable for the task of measuring soft skills. Second, we combine expert consensus forecast and machine learning to better predict future job trends. Mixed methods which complement data-driven forecasting techniques with the insight of teams of experts are increasingly utilized to better predict macroeconomic trends. We contribute to this literature by developing and implementing a new strategy to integrate qualitative (human) and quantitative (machine) forecasting capabilities. Finally, we derive a series of policy implications for R&D, education, productivity and employment. This project contributes to address the socioeconomic effects of globalization and technological transformations by harnessing the combined power of human and artificial intelligence.

P.I.: Massimo Riccaboni

PRIN 2017

Cognitive Modes, Social Motives and Prosocial Behavior

This project brings together a variety of competences and methods to address an issue that has recently gained attention by scholars working at the intersection of different disciplines: economics, psychology, sociology, cognitive sciences. The issue under debate concerns the interaction between cognitive modes and prosocial behavior, i.e., the extent to which individual behavior is beneficial to the society as a whole. The basic question is: which cognitive mode is more likely to foster prosocial behavior? Up to now, such a debate is lively and far from being set. Indeed, the literature provides both conflicting theoretical arguments and inconclusive empirical evidence 

P.I.: Ennio Bilancini


Studio avanzato delle reti sociali di collaborazione e di innovazione tramite lo strumento di Social Network

Il progetto prevede lo studio avanzato delle reti sociali, di collaborazione e di innovazione emergenti in contesti sociali e organizzativi tramite lo strumento di Social Network Analysis. L'attività di ricerca è volta alla finalizzazione di una metodologia avanzata di Network Analysis nonché alla valorizzazione della qualità all'interno del Knowledge Management System (#KMS).

P.I.: Massimo Riccaboni

Made in viareggio

Il Made in Viareggio nell'industria nautica: pattern recognition, tecnologie abilitanti e competenze per la digital innovation

La ricerca focalizza e inquadra il modello di business e il sistema valoriale di riferimento, e al suo interno il ruolo delle competenze e delle risorse umane, nelle imprese manifatturiere operanti nel distretto della nautica e della portualità toscana avendo riguardo all’introduzione delle tecnologie abilitanti note come “Industria 4.0” e al processo di digital innovation.

L’obiettivo del progetto è la stesura di una policy design strategy che (a) raccolga e racchiuda lo stato dell’arte della situazione mediante la somministrazione di un survey e che (b) indichi come le tecnologie abilitanti possano contribuire al processo di digital innovation avendo il focus sul ruolo delle competenze e delle risorse umane e che (c) disegni percorsi specifici di formazione e business education e attivi insieme al partner selezionato (Consorzio Na.Vi.Go. s.c. a r.l., da ora in avanti Na.Vi.Go., la più estesa rete di aziende di nautica da diporto della Toscana, con finalità di promozione e coordinamento delle attività imprenditoriali del distretto nautico di Viareggio) le iniziative più idonee al trasferimento tecnologico e alla strutturazione delle competenze qualificanti e abilitanti.

P.I.: Mirco Tribastone