CEC | Introduction of artificial intelligence in the PA workflow management software system in response to citizens' requests
The goal of the project concerns the creation and testing of a Service Layer composed of a series of Artificial Intelligence and Natural Language Processing services, which provides the end customer with a framework for configuration and guided adaptation activities to their own domain through a simple administration dashboard.
Starting from this, you will experience the possibility of:
- identify significant entities and information nuclei within the reference domain
- extract topics and create domain ontologies semi-automatically
- apply text summarization algorithms to texts
- extract elements of similarity to feed text analysis engine
The project will aim to lay the foundations for innovative domain adaptation and linguistic semantic analysis services in SaaS mode, through a Service Layer that focuses in particular on two specific pilot cases that allow demonstrating the feasibility of a wider robotic process initiative automation for the PA:
1) Analysis of the documents entering the protocol to carry out a (possibly) automatic redirection to the specific service on the basis of a standard file.
2) Suggestion of "collating" activities of documents contained in a repository, which are consistent for administrative procedure and instance. |
CIRCLE GROUP
DM 117 | Advanced approaches for multi-robot coordination
Robotic technology has recently matured enough to implement autonomous robotic systems for everyday use in diverse applications: from disaster response to environmental monitoring and logistics. In such applications, robots must establish collaborative interactions so as to achieve their individual and collective goals and a key problem is that robots make individual decisions so as to optimize an objective function of the whole system. This problem is typically referred to as coordination.
Robotic technology has recently matured enough to implement autonomous robotic systems for everyday use in diverse applications: from disaster response to environmental monitoring and logistics. In such applications, robots must establish collaborative interactions so as to achieve their individual and collective goals and a key problem is that robots make individual decisions so as to optimize an objective function of the whole system. This problem is typically referred to as coordination.
This PhD thesis will examine modern optimization techniques for coordination in multi-robot systems implemented in a logistics application such as the transport of items from loading docks to unloading docks. The approach will be quantitatively evaluated in a simulated scenario using standard robotic tools (ROS and Gazebo). |
CIRCLE GROUP
DM 117 | Efficiency techniques for analyzing speech signals for intelligent support agents
The interpretation of the voice and its consequent understanding, today represents a considerable area of interest in numerous areas of the technology offered on the market both in the industrial and domestic sectors. This propensity for the development of voice-based technologies is given by the obviousness of the naturalness of the vocal means of communication, which can bring considerable benefits where it is not possible to use other communication tools. Furthermore, it is undeniable that the evolution of Natural Language Processing (NLP) tools requires constant evolution in the interpretation of data not only in an analytical but conscious and proactive way.
These considerations, combined with the lack of truly effective and robust tools, lead to the need to further explore this particular branch of Artificial Intelligence in order to be able to improve what is currently on the market.
The aim of the research activity covered by the doctorate is, therefore, to study and implement innovative techniques for processing vocal signals in order to build an artificial intelligence capable of providing assistance in real time effectively and efficiently.
In particular, through the use of Machine Learning and the most advanced Artificial Intelligence techniques, we hope to create an autonomous software agent capable of being used:
- in the industrial field, as a control room assistant or in case of maintenance where the operator is unable to use his hands and
- in the domestic environment, able to provide a greater level of confidence to the end user, a particularly sensitive use case in the field of aging and motor disabilities. The research and development path in question was agreed with Hitachi Rail in order to create a synergy of intent in the NLP field for industrial fields.
The research activity will be carried out in collaboration with the Circle Garage company (https://circlegarage.com/). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences. |
ENGINEERING | Implementation of a Decision Support System for the efficiency of the management of an integrated Water Management System, Engineering S.p.A.
Development of a DSS (decision support system) for the management of an integrated water system that includes predictive analysis based on machine learning and AI. The system will have to implement best practices for the control and management of hydraulic efficiency (e.g. District Metered Area) and quantify efficiency based on standard indicators such as IWA, UKWIR for the analysis of water losses. A prototype development will be envisaged within the system envisaged in this project to implement hydraulic modeling functions (Digital Twin) useful for performing the routine operation of the network, allowing operators to identify better operating scenarios for the plants and the network, to improve the proactive management of critical events and to predict the effects of planned operations or anomalies.
Implementazione di un Sistema di Supporto alle Decisioni per l'efficienza della gestione di un Sistema Integrato di Gestione delle Acque, Engineering S.p.A.
Sviluppo di un DSS (decision support system) per la gestione di un sistema idrico integrato che includa analisi predittive basate su machine learning e AI. Il sistema dovrà implementare le migliori pratiche per il controllo e la gestione dell'efficienza idraulica (es. District Metered Area) e quantificare l'efficienza sulla base di indicatori standard come IWA, UKWIR per l'analisi delle perdite idriche. All'interno del sistema previsto in questo progetto sarà previsto uno sviluppo prototipale per implementare funzioni di modellazione idraulica (Digital Twin) utili a svolgere il funzionamento di routine della rete, consentendo agli operatori di identificare migliori scenari operativi per gli impianti e la rete, per migliorare la gestione degli eventi critici e per prevedere gli effetti di operazioni programmate o anomalie. |
ESAOTE | Machine Learning for Ultrasound Imaging
The focus of this 3 years position is the application of machine learning algorithms, libraries and tool for the analysis of ultrasound images. The research activity will be carried out in collaboration with the Esaote company (https://www.esaote.com) and the MaLGa and Computer Vision research groups of Dibris (Prof. Francesca Odone). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
Apprendimento automatico per analisi di immagini ad ultrasuoni
Il focus di questa posizione triennale è l'applicazione di algoritmi di apprendimento automatico, librerie e strumenti per l'analisi di immagini ecografiche. L'attività di ricerca sarà svolta in collaborazione con la società Esaote (https://www.esaote.com) e i gruppi di ricerca MaLGa e Computer Vision del Dibris (Prof. Francesca Odone). I candidati svolgeranno un programma di dottorato in stretta collaborazione tra università e industria e saranno supportati da un supervisore accademico e da un supervisore del dipartimento di ricerca e sviluppo dell'azienda. I candidati devono avere una solida preparazione in informatica. Il Dottorato prevede attività di formazione su temi avanzati, attività di ricerca e trasferimento tecnologico su temi di interesse per il dipartimento R&D aziendale e per il supervisore accademico. Durante il percorso di dottorato, i candidati avranno l'opportunità di partecipare a seminari, workshop e scuole di dottorato organizzati dal Collegio dei docenti del CSSE e workshop e conferenze nazionali e internazionali. |
IRCCS GASLINI | L'Istituto Gaslini nel suo percorso di digitalizzazione complessiva e di integrazione dei sistemi, supporto della clinica e della ricerca, intende aumentare la qualità, la condivisione ed il contenuto informativo del dato (piramide DIKW) con progetti che portino all'effettiva adozione dei più recenti standard di codifica (ontologie) e di rappresentazione dei dati clinici (es.:FHIR), di definizione dell'architettura e delle metodologie di analisi dei dati (es.: data mining), di business reengineering, di evoluzione dei sistemi di business intelligence e di IA per l'analisi del linguaggio e l'estrazione dei contenuti informativi.
L'attività di ricerca sarà svolta in collaborazione con l'Istituto Gaslini e (https://www.gaslini.org/). I candidati svolgeranno un programma di dottorato in stretta collaborazione tra università e istituto e saranno supportati da un supervisore accademico e da un supervisore del dipartimento di ricerca e sviluppo dell'azienda. I candidati devono avere una solida preparazione in ambito informatico. Il corso di dottorato prevede attività di formazione su tematiche avanzate, attività di ricerca e trasferimento tecnologico su temi di interesse per il dipartimento R&S aziendale e per il tutor accademico. Durante il programma di dottorato, i candidati avranno l'opportunità di frequentare seminari, workshop e scuole di dottorato organizzati dal collegio di dottorato CSSE e workshop e conferenze nazionali e internazionali. |
GTER | GNSS precise and affordable positioning in harsch environment
The project aims to analyze and investigate the potential of new services and new devices for satellite positioning.
Mass market GNSS receivers, together with the availability of antennas with new technologies and the availability of new augmentation services (eg: Galileo HAS), open up a series of new applications.
In particular in an urban environment and in non-nominal conditions for GNSS positioning, we want to test new online and offline processing algorithms that allow us to guarantee robust positioning in any case.
This becomes a necessary condition to extend the use of GNSS also to applications, some safety critical, where precision and accuracy are required. Think of everything that revolves around autonomous driving, but also of applications related to navigation in urban environments or precision tracking.
Posizionamento GNSS efficiente e di precisione in ambienti difficili
Il progetto si propone di analizzare e indagare le potenzialità di nuovi servizi e nuovi dispositivi per il posizionamento satellitare.
I ricevitori GNSS del mercato di massa, insieme alla disponibilità di antenne con nuove tecnologie e alla disponibilità di nuovi servizi di potenziamento (es: Galileo HAS), aprono una serie di nuove applicazioni.
In particolare in ambiente urbano e in condizioni non nominali per il posizionamento GNSS, vogliamo testare nuovi algoritmi di elaborazione online e offline che ci permettano di garantire comunque un posizionamento robusto.
Questa diventa una condizione necessaria per estendere l'uso del GNSS anche ad applicazioni, alcune safety critical, dove sono richieste precisione e accuratezza. Si pensi a tutto ciò che ruota intorno alla guida autonoma, ma anche alle applicazioni legate alla navigazione in ambienti urbani o al tracciamento di precisione. |
LEONARDO Regione Liguria | Trustworthy AI applied to certification AI System
This project aims at developing provably trustworthy and efficient algorithms for modern machine learning applications. While often achieving impressive results, state of the art AI systems can be brittle and inefficient, in this projects we will work towards developing novel compressed algorithms coming with rigorous guarantees in terms of accuracy and computational costs. The ideal candidate should have excellent computational and algorithmic skills, while being keen on engaging both with theoretical and practical aspects.
The research activity will be carried out in collaboration with the Leonardo company (https://www.leonardo.it/) in their Labs (AI, HPC, Big Data). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
Affidabile AI applicata alla certificazione AI System
Questo progetto mira a sviluppare algoritmi affidabili ed efficienti per le moderne applicazioni di machine learning. Pur ottenendo spesso risultati impressionanti, i sistemi di intelligenza artificiale all'avanguardia possono essere fragili e inefficienti, in questo progetto lavoreremo allo sviluppo di nuovi algoritmi compressi dotati di rigorose garanzie in termini di accuratezza e costi computazionali. Il candidato ideale dovrebbe avere eccellenti capacità computazionali e algoritmiche, pur essendo desideroso di impegnarsi sia con aspetti teorici che pratici.
L'attività di ricerca sarà svolta in collaborazione con la società Leonardo (https://www.leonardo.it/) nei loro Laboratori (AI, HPC, Big Data). I candidati svolgeranno un programma di dottorato in stretta collaborazione tra università e industria e saranno supportati da un supervisore accademico e da un supervisore del dipartimento di ricerca e sviluppo dell'azienda. I candidati devono avere una solida preparazione in ambito informatico. Il corso di dottorato prevede attività di formazione su tematiche avanzate, attività di ricerca e trasferimento tecnologico su temi di interesse per il dipartimento R&S aziendale e per il tutor accademico. Durante il programma di dottorato, i candidati avranno l'opportunità di frequentare seminari, workshop e scuole di dottorato organizzati dal collegio di dottorato CSSE e workshop e conferenze nazionali e internazionali. |
LEONARDO DM 117 | Parallel batch processing for Trino workflows on GPUs.
Essentially, a new Trino branch should be created while completing the implementation of core data processing capability functions, with hardware independent GPUs back end. Such development shall be accomplished through Domain Specific Language (DLS) approach and abstraction. Here is the official git repo issue: https://github.com/trinodb/trino/discussions/13729 . The success KPI for the activity would be acceptance of the pull request to the main repository by Trino developers committee and Journal Paper grade publications. The research activity will be carried out in collaboration with the Leonardo company (https://www.leonardo.it/) in their Labs (AI, HPC, Big Data). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
|
LEONARDO DM 117 | A Detailed study toward adaptive optimization of data intensive workflows based on Trino and Spark.
A Detailed study toward adaptive optimization of data intensive workflows based on Trino and Spark, in terms of interaction of the system calls at the OS level in case of different type of virtualization approaches, such as Singularity, Docker and OpenStack VMs. A complete benchmarking shall be assessed in terms of gRPC optimization and reduction of computational latencies also at OS scheduler level and I/O systems throughput (Object Strage and Parallel File System). A data driven adaptive approach shall be developed for assessing online computational efficiency optimization during batch processing. The research activity will be carried out in collaboration with the Leonardo company (https://www.leonardo.it/) in their Labs (AI, HPC, Big Data). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
|
PIRELLI DM 117 | The research activity will be carried out in collaboration with the Pirelli Type R&D labs (https://www.pirelli.it/). Machine learning methods can be greatly enhanced when a priori information on the problem at hand can be incorporated in the algorithms design. In this project, we will consider so called physics informed approaches that has been recently shown to provide impressive results. The goal will be to study and develop sound and efficient algorithms. In particular, we will investigate these methods in the context of rubber compounds in collaboration with Pirelli. The incorporation of physics informed machine learning techniques will complement and support traditional development approaches in rubber design, facilitating the development of novel compounds and optimizing the overall design process. Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
L'attività di ricerca sarà svolta in collaborazione con i laboratori Pirelli Type R&D (https://www.pirelli.it/). I candidati svolgeranno un programma di dottorato in stretta collaborazione tra università e industria e saranno supportati da un supervisore accademico e da un supervisore del dipartimento di ricerca e sviluppo dell'azienda. I candidati devono avere una solida preparazione in ambito informatico. Il corso di dottorato prevede attività di formazione su tematiche avanzate, attività di ricerca e trasferimento tecnologico su temi di interesse per il dipartimento R&S aziendale e per il tutor accademico. Durante il programma di dottorato, i candidati avranno l'opportunità di frequentare seminari, workshop e scuole di dottorato organizzati dal collegio di dottorato CSSE e workshop e conferenze nazionali e internazionali. |
SNOOTS DM 117 | The aim of the project is to study, develop and validate computer vision and deep learning methods for the analysis of medical images in the veterinary field. The results of the project will be used to create a new generation of applications capable of helping the veterinarian to make better diagnoses thanks to automatic tools and the systematic application of best practices. The project will mainly focus on radiology but other acquisition modalities, such as ultrasound images, will also be considered. Machine learning tasks to be studied will include artifact and noise removal, image enhancement, image classification, and automatic detection of anatomical measurements.
The research activity will be carried out in collaboration with the SNOOTS company (https://www.snoots.pet/). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
L'attività di ricerca sarà svolta in collaborazione con la società SNOOTS (https://www.snoots.pet/). I candidati svolgeranno un programma di dottorato in stretta collaborazione tra università e industria e saranno supportati da un supervisore accademico e da un supervisore del dipartimento di ricerca e sviluppo dell'azienda. I candidati devono avere una solida preparazione in ambito informatico. Il corso di dottorato prevede attività di formazione su tematiche avanzate, attività di ricerca e trasferimento tecnologico su temi di interesse per il dipartimento R&S aziendale e per il tutor accademico. Durante il programma di dottorato, i candidati avranno l'opportunità di frequentare seminari, workshop e scuole di dottorato organizzati dal collegio di dottorato CSSE e workshop e conferenze nazionali e internazionali. |
| |
| |
| |
MANYDESIGN Reserved for
Manydesign
Employees | Definition of a model for semi-automatic creation and updating of skills and competences inside a Company
A problem of large companies is the knowledge of corporate skills, made up of the knowledge of individual people. This knowledge can be obtained at the time of hiring, but then changes over time depending on the person’s experience within the projects, courses and certifications obtained. The PhD project will deal with defining a skills model that allows to follow the evolution of skills, to be able to be queried to find skills within the company and to highlight the strengths of the company itself . There will be a study to define usable dashboard in interactive ways.It will also be defined how to populate the model automatically by analysing in a Software Company the projects within the Git repositories, the static code analysis systems (to understand security skills), issue tracking systems, other systems, document (like cv), etc...Strong importance will be given to the visibility and privacy of data by integrating with the company organization chart.
The research activity will be carried out in collaboration with the Manydesign company (https://manydesigns.com). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences. |
TECHNOLEADER
COBO GROUP DM 117 | IT security of interconnected mobile devices
The research activity will be carried out in collaboration with COBO Engineering and Technoleader (https://www.technoleader.com/). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
Sicurezza informatica dei dispositivi mobili interconnessi
L'attività di ricerca sarà svolta in collaborazione con COBO Engineering e Technoleader (https://www.technoleader.com/). I candidati svolgeranno un programma di dottorato in stretta collaborazione tra università e industria e saranno supportati da un supervisore accademico e da un supervisore del dipartimento di ricerca e sviluppo dell'azienda. I candidati devono avere una solida preparazione in ambito informatico. Il corso di dottorato prevede attività di formazione su tematiche avanzate, attività di ricerca e trasferimento tecnologico su temi di interesse per il dipartimento R&S aziendale e per il tutor accademico. Durante il programma di dottorato, i candidati avranno l'opportunità di frequentare seminari, workshop e scuole di dottorato organizzati dal collegio di dottorato CSSE e workshop e conferenze nazionali e internazionali. |
TECHNOLEADER
COBO GROUP DM 117 | Machine safety, mobile vehicle automation, and runtime verification.
The research activity will be carried out in collaboration with COBO Engineering and Technoleader (https://www.technoleader.com/). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
Sicurezza delle macchine, automazione dei veicoli mobili e verifica del runtime.
L'attività di ricerca sarà svolta in collaborazione con COBO Engineering e Technoleader (https://www.technoleader.com/). I candidati svolgeranno un programma di dottorato in stretta collaborazione tra università e industria e saranno supportati da un supervisore accademico e da un supervisore del dipartimento di ricerca e sviluppo dell'azienda. I candidati devono avere una solida preparazione in ambito informatico. Il corso di dottorato prevede attività di formazione su tematiche avanzate, attività di ricerca e trasferimento tecnologico su temi di interesse per il dipartimento R&S aziendale e per il tutor accademico. Durante il programma di dottorato, i candidati avranno l'opportunità di frequentare seminari, workshop e scuole di dottorato organizzati dal collegio di dottorato CSSE e workshop e conferenze nazionali e internazionali. |
TECHNOLEADER COBO GROUP DM 117 | Use of cameras for security purposes and/or automation of mobile vehicles
The research activity will be carried out in collaboration with COBO Engineering and Technoleader (https://www.technoleader.com/). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
Utilizzo di telecamere per finalità di sicurezza e/o automazione di mezzi mobili
L'attività di ricerca sarà svolta in collaborazione con COBO Engineering e Technoleader (https://www.technoleader.com/). I candidati svolgeranno un programma di dottorato in stretta collaborazione tra università e industria e saranno supportati da un supervisore accademico e da un supervisore del dipartimento di ricerca e sviluppo dell'azienda. I candidati devono avere una solida preparazione in ambito informatico. Il corso di dottorato prevede attività di formazione su tematiche avanzate, attività di ricerca e trasferimento tecnologico su temi di interesse per il dipartimento R&S aziendale e per il tutor accademico. Durante il programma di dottorato, i candidati avranno l'opportunità di frequentare seminari, workshop e scuole di dottorato organizzati dal collegio di dottorato CSSE e workshop e conferenze nazionali e internazionali. |
TEMA srl DM 117 | The research activity will be carried out in collaboration with Tema srl (https://temasrl.net/). Candidates will carry out a doctoral program in close collaboration between university and industry and they will be supported by an academic supervisor and a supervisor from the company's R&D department. Candidates are required to have a solid background in computer science. The PhD program includes training activities on advanced topics, research and technology transfer activities on themes of interest for the company R&D department and for the academic supervisor. During the PhD program, candidates will have the opportunity to attend PhD seminars, workshops, and schools organized by the CSSE PhD board and national and international workshops and conferences.
|
