Reserved topic scholarships and PhD Executive positions

The Doctorate Program in Industrial Innovation offers mainly 2 types of 3-year PhD positions:

• Standard PhD positions with scholarships financed by the companies;
• PhD Executive positions, which are without scholarships and reserved to the partner companies' employees who maintain their status (and salary) within their company. The applicants can be either already employed by the company, or they can be employed after the selection process, but before the PhD Program starts. The financial conditions are defined directly between the company and the applicant.

41st Cycle - Intake year 2025

Adige S.p.A. - 1 scholarship

A - AI for internal technical office process management (Project Management, Generation of user manuals documents or classification of material master data)

Product development faces numerous challenges, including inefficiencies, competitors, and political instability. This PhD research proposal aims to develop a comprehensive model that monitors the progress of product development projects and that realizes and analyzes possible project portfolio scenarios based on resource load as the environment around us evolves.  This is to have a tool that allows us to improve decision-making and operational efficiency. 
The primary objectives of this research are threefold: firstly, to develop a robust model that allows the simulation of project portfolio scenarios; secondly, integrate systems to provide continuous monitoring and feedback; thirdly, to implement advanced artificial intelligence and machine learning algorithms to analyze this data and suggest optimizations to the portfolio itself and suggest the advantages of one over another; 
The methodology involves an extensive literature review, followed by the design and implementation of the simulation model and environment. AI models will be developed and trained for predictive analytics and optimization, with simulations conducted to test the effectiveness of the system. This research is expected to have a significant impact on managing the product development process by improving efficiency, reducing costs, and improving customer satisfaction. In addition, it will contribute academically by advancing knowledge in technology and optimizing the use of emerging technologies applied in the management field. 

Required/Preferred Candidate Skills and Competencies:
Candidates for this research should possess a strong foundation in computer science, with expertise in artificial intelligence and machine learning. Essential skills include data integration, real-time systems, and software development, with proficiency in programming languages. Understanding the principles of project management. In addition, candidates must demonstrate strong analytical and problem-solving skills, capable of designing and implementing complex systems. In addition to technical skills, curiosity and a passion for innovation are key. Successful candidates will be driven by a desire to explore new ideas, ask in-depth questions, and pursue new solutions, while continuously learning and adapting to emerging trends in the field.

Contact: Giovanni Iacca (UniTrento), Marco Formentini (UniTrento), Manuela Valentinotti  (BLM group - Adige)

UniTrento in collaboration with CIRA Centro Italiano Ricerche Aerospaziali SCpA - 1 scholarship

B -  Structural optimization of satellite-sensor interfaces for geodesic detection systems (Project "SPACE IT UP! Contratto ASI n.2024-5-E.0 Spoke 6, CUP Master n.  I53D24000060005”, CUP di progetto n.  E63C24000530003)

The research focuses on the development of enabling technologies for the mapping of Earth’s gravitational field using Gravitational Reference Systems placed in Earth orbit. These systems are designed to generate inertial references with minimal disturbance, which are essential for providing accurate signals of the gravitational field. The mechanical interface between sensors and the satellite platform is critical to ensure measurement accuracy, as structural vibrations, thermal distortions, or misalignments can introduce noise and degrade data quality. 

The research project addresses the problem of lightweight design and the testing of subsystems that make up the gravitational reference system. In particular, this project aims to design, model, and experimentally validate advanced structural interfaces between satellite payloads and geodetic sensors. The structural design of the critical components of the system requires an optimization procedure that within the design constraints identifies the optimal stiffness and strength conditions, to be implemented and verified on the prototype. 

The study will also integrate smart materials in innovative configurations and structures to exploit their properties for dynamic load mitigation and precision alignment.

The expected outcome is the design optimized interfaces enabling higher precision in geodetic measurements by minimizing mechanical noise. The research will also contribute to the broader field of adaptive aerospace structures, with potential spin-offs for deployable space mechanisms or reconfigurable satellite components.

Required/Preferred Candidate Skills and Competencies:

Master’s degrees in Mechanics or Mechatronics Engineering; Materials Engineering; Aerospace Engineering; Civil Engineering, Physics awarded by an Italian university or a title obtained abroad which has been recognized as equivalent to the Italian degree.

Competencies are required in structural design, mechanics of vibrations, finite element modeling, materials and application of structural optimization techniques.

The candidate is expected to have interest and skills in the analysis of aerospace structures and orbiting Earth observation systems.

Contact: Vigilio Fontanari (UniTrento), Emiliano Rustighi (UniTrento),  Antonio Concilio  (CIRA)

FBK in collaboration with Sony Europe Limited – Italian Branch - 2 scholarships

C - Design of a highly efficient sensor-embedded machine learning accelerator for imaging systems 

Sensing devices interact in complex environments, and their miniaturization and portability call for small form factor and low power consumption. Edge AI is therefore of fundamental importance for robustness, privacy, and long battery operation. The application of AI algorithms to image sensors further constraints technological solutions in resources due to the large amount of generated data. 
The objective of this project is to design and prototype a digital circuit that implements a parallel HW/SW architecture aimed at solving specific sensing use cases with high efficiency in area and power. The candidate will tackle the challenges with a multidisciplinary point of view, focusing on the analysis of a proposed HW/SW architecture and the design of a digital prototype.
The key technical activities that will be covered are:
●    Implementation of a prototype design in a mixed hardware/software platform (which might include PC, SoC, FPGAs, integrated circuits).
●    Simulations of the system behavior in target scenarios.
●    Characterization of the developed system in the lab and on the field.
The candidate will meet regularly with an industrial advisor from Sony and will spend a research period of six months at the company, providing a unique opportunity to interact with experts in the fields of computer vision and CMOS image sensors. The expected outcome is a highly optimized architecture for an imaging system that combines software, firmware, and hardware solutions to solve specific, market-driven needs in the consumer electronics, automotive, smart city, robotics, and digital industry fields. The intellectual property of the research results that will derive from the activities carried out by the doctoral candidate is owned by FBK and the Company.

Required/Preferred Candidate Skills and Competencies:

- Proficiency in digital electronics and FPGA/IC design.
- Familiarity with computer vision and deep learning.
- Knowledge of image sensor architectures.
- Software programming using C++, Python, Matlab.
- Testing of electronic devices and systems.

Contact: Leonardo Gasparini (FBK), Davide Marani (Sony)

D - Towards disruptive IR imaging with antenna-coupled field-effect transistor detectors

This project aims to expand portable device capabilities beyond human vision through advanced THz to thermal infrared sensing and imaging systems. Following a structured research, the work begins with comprehensive study and modeling of how to bring antenna-coupled field-effect transistors can detect THz radiation and be optimized for the thermal infrared region exploiting the plasma-wave principle. After establishing theoretical foundations, the student will design and optimize individual sensing structures specifically tailored for infrared detection, validating their performance through rigorous characterization. The final phase will focus on scaling these optimized single-pixel designs into a complete infrared imaging array with appropriate readout and control circuitry. This multidisciplinary approach requires expertise in quasioptical and electromagnetic principles, device physics, noise analysis, and integrated circuit design. Technical work encompasses high-level modeling of devices, sensitivity optimization, noise propagation analysis, novel pixel architecture development, and full circuit implementation from schematic to layout. 
The candidate will meet regularly with an industrial advisor from Sony and will spend a research period of six months at the company, providing a unique opportunity to interact with experts in the fields of computer vision and CMOS image sensors.
Through collaboration with experts in image sensors, THz/IR detectors, and analog circuit design, the student will develop specialized knowledge across these domains. The project objective is the demonstration of a functioning infrared imaging array. The intellectual property resulting from the doctoral student's research activities will be owned by FBK and the Company.

Required/Preferred Candidate Skills and Competencies:

- Background in analog electronics and mixed signal IC design
- Knowledge of image sensor architectures
- Antenna design
- Testing of electronic devices and systems
- Software programming for the generation of high-level models

Contact: Leonardo Gasparini (FBK), Daniele Perenzoni (Sony)

VEASYT srl - 1 additional scholarship (published on 29/07/2025)

E - Architectures and Models for Efficient Sign Language Video Generation

This PhD project will explore generative AI techniques for the automatic creation of realistic and intelligible Italian Sign Language (LIS) videos. The research will focus on developing and evaluating models capable of producing real-time photo-realistic interpreter animations that are semantically faithful to the source text, in order to meet both linguistic and accessibility requirements.

Required/Preferred Candidate Skills and Competencies:
- Familiarity with recent advances in computer vision and deep learning (e.g., Vision Transformers, Diffusion Models, Foundation Models).
- Ability to read and implement ideas from research papers (e.g., NeurIPS, CVPR, ICCV).
- Proficient in Python

Contact: Elisa Ricci (UniTrento), Ivan Grasso (VEASYT srl)

Electrolux Italian SpA - 1 PhD Executive

AE - Embedded Computing: household appliances application case

The goal of this project is to make a step towards smart home appliances that autonomously improve their performance over time based on experience gained from different situations in the consumer’s home. It is envisioned that to have a big leap in this direction, it is a prerogative to run real-time applications locally on the embedded system itself, without the use of cloud servers, overcoming some of the challenges such as latency, privacy, scalability, and sustainability. The main objective of this project is, therefore, to consolidate an end-to-end workflow that facilitates the development of a secure, robust solution from idea through integration into embedded systems. This pipeline shall be beneficial in speeding up the deployment to market for any similar application. A specific representative and high-impact use case shall be chosen to develop state-of-the-art, data-driven control algorithms that are capable of adapting to the changing dynamics of the environment and natural degradation of components, and tuned, where possible, to match user needs. The main technologies that will be considered are Reinforcement Learning, Online learning, and Embedded AI. As for the MLOps, Profiling and Automatic Code Generation custom pipelines shall be developed, to avoid depending on specific HW vendors. 

Required/Preferred Candidate Skills and Competencies:

The candidate shall demonstrate a strong background in Embedded AI with the final vision to exploit it in the development of smart functionalities for IOT devices, mainly devoted to the white goods sector. Considering the same sector or equivalent ones, the candidate shall demonstrate a well-structured knowledge about technologies like Reinforcement Learning and Online learning. Experience in full-stack development and validation of AI models is mostly appreciated, considering the need for the deployment process on embedded platforms. The candidate shall demonstrate fluency in English, written and spoken, and have a good capability of bridging between the academic environment and the industrial one. 

The intellectual property of the research results that will derive from the activities carried out by the doctoral student is owned by the Company.

Contact: Iacca Giovanni (UniTrento), Gilberto Pin (Electrolux)

Huawei Technologies Duesseldorf GmbH - 2 PhD Executive 

BE - AI system security

The PhD candidate will contribute to the research, development, publication and product adoption of system security technologies for next-generation AI and HPC infrastructure leveraging heterogeneous computing clusters based on CPUs, NPUs and GPUs, working on real world problems for a wide range of Huawei products and services to preserve and prove their trustworthiness through technology, from a unique research position connecting Academia and Industry. Potential responsibilities include: 
• Assess system security concerns for AI and HPC computations running on xPU (e.g. NPU, GPU, TPU) as part of a heterogeneous computing clusters 
• Research on innovative concepts and architectures in the field of Trustworthy Computing
• Implement proof of concepts and support technology adoption in products
• Publish research papers and technical reports at top security conferences and collaborate with academic partners
• Improve, by design, the end to end system security capabilities of future Huawei products

Required/Preferred Candidate Skills and Competencies:
• You have recently completed or will soon complete your master studies in computer science, information technology, cybersecurity, electrical engineering, or mathematics
• You have research experience in system security, applied cryptography, computing engines such as GPU or NPU, or distributed computing
• Experience in one or more of the following is a plus: compilers, AI framework design (full stack), memory safety, TEEs
• You have good programming skills in Rust, C, C++, Go and in system programming
• You have demonstrated affinity for concept design and hands-on validation
• You have a passion for finding solutions for complex technology issues
• You have excellent collaboration and communication skills

The intellectual property of the research results that will derive from the activities carried out by the doctoral student is owned by the Company.

Contact: Bruno Crispo (UniTrento), Silviu Vlasceanu (Huawei Technologies Duesseldorf Gmbh)

CE - System security for next-generation computing

The PhD candidate will contribute to the research, development, publication and real-world deployment of system security technologies for next-generation computing platforms, with a particular focus on Neural Processing Units (NPUs) and Graphics Processing Units (GPUs) with respect to their evolving roles in accelerating the development of agentic large language models (LLMs). This research position will explore security challenges and opportunities at the intersection of hardware and software in contemporary computing platforms comprising CPUs, NPUs, and GPUs.

The position offers a unique opportunity to work on cutting-edge problems that impact the trustworthiness of current and future computing systems, bridging academic research with practical deployment across Huawei’s product ecosystem. Potential responsibilities include: 

• Research and address system security challenges in evolving computing platforms, particularly focusing on integrated accelerators, i.e., NPUs and GPUs, and their role in securely executing data-intensive applications, such as LLMs. 
• Explore new attack vectors on these computing platforms, and research novel defenses accordingly. 
• Implement proof of concepts and support technology adoption in products
• Publish research papers and technical reports at top security conferences and collaborate with academic partners
• Improve, by design, the end to end system security capabilities of future Huawei products

Required/Preferred Candidate Skills and Competencies:

• You have recently completed or will soon complete your master studies in computer science, information technology, cybersecurity, electrical engineering, or mathematics
• You have research experience in system security, applied cryptography, computing engines such as GPU or NPU, or distributed computing
• Experience in one or more of the following is a plus: compilers, AI framework design (full stack), memory safety, TEEs
• You have good programming skills in Rust, C, C++, Go and in system programming
• You have demonstrated affinity for concept design and hands-on validation
• You have a passion for finding solutions for complex technology issues
• You have excellent collaboration and communication skills

The intellectual property of the research results that will derive from the activities carried out by the doctoral student is owned by the Company.

Contact: Bruno Crispo (UniTrento), Silviu Vlasceanu (Huawei Technologies Duesseldorf Gmbh)

FBK- 3 PhD Executive 

DE - Next-Generation Agri-Robotics: Merging Perception, Control, and Autonomy for Minimal-Interaction Farming

This research will focus on developing autonomous, cost-effective systems at the intersection of computer vision, robotics, and predictive control for precision agriculture. These systems are intended to operate reliably in real-world farming environments with minimal human intervention. The work will target key agricultural tasks such as crop monitoring, targeted weeding, harvesting, and resource management through intelligent, adaptive technologies. By integrating advanced perception, control, and decision-making capabilities, the project aims to enhance operational efficiency, reduce labor requirements, and optimize resource usage. This is not foundational research; rather, it is applied research with a strong focus on advancing the state-of-the-art in autonomous rover platforms for agriculture. The expected outcome is the development of practical, scalable, and robust solutions that contribute to sustainable and resilient farming systems.

Required/Preferred Candidate Skills and Competencies:
Required:
- Strong background in robotics, computer vision, or control systems;
- Proficiency in programming (e.g., Python, C++, or Java);
- Proficiency in version control systems and methodologies (e.g., Git);
- Familiarity with machine learning principles or fast prototyping of embedded systems;
- Inclination to openly share ideas and development directions with team members;
- Ability to work independently and in a multicultural team;
- Good written and verbal communication skills;
Preferred:
- Proven experience with ROS/ROS2;
- Proven experience with autonomous systems and/or agricultural technologies;
- Proven experience in research paper writing and scientific communication;
- English language certification.

The intellectual property of the research results that will derive from the activities carried out by the doctoral student is owned by the Company.

Contact: Massimo Vecchio (FBK)

EE - Advanced model tool (AI-based, ROM, ML, etc) for improving performance on electrolysis cell and stack

The student will be involved in activities to improve PEM electrolysers. 

Activities can include:
● Physics-informed machine learning for model order reduction, degradation prediction, and control
● EIS analysis of electrolysis cells and development of (AI-based)  models for optimising cell components and operating conditions
● Enhancing Performance of Low-Temperature Electrolysers via Computational Fluid Dynamics Analysis

Required Candidate Skills and Competencies: 
- Strong background in physics or engineering
- Familiarity with hydrogen and gas/solid reaction
- Ability to work independently and in a multicultural team;
- Good written and verbal communication skills;

The intellectual property of the research results that will derive from the activities carried out by the doctoral student is owned by the Company.

Contact: Matteo Testi (FBK)

FE - Metal hydride application: novel tank design and thermochemical compression

The student will be involved the activities on metal hydride technologies, in particular:

● Metal hydride as solution for thermal compression/purification with nove tank design. 
● Entropy Generation Analysis for Optimizing the Design and Operation of Hydrogen Solid-State Storage Systems

Required/Preferred Candidate Skills and Competencies:Required:
- Strong background in physics or engineering
- Familiarity with hydrogen and gas/solid reaction
- Ability to work independently and in a multicultural team;
- Good written and verbal communication skills.

The intellectual property of the research results that will derive from the activities carried out by the doctoral student is owned by the Company.

Contact: Matteo Testi (FBK)

Ubitech - 1 PhD Executive 

GE - Towards Trustworthy Federation of AI Agents leveraging HW-based Trusted Computing Abstractions

The deployment of agentic AI systems – based on transformer architectures – in industrial settings poses several critical challenges. Ensuring reliable and trustworthy AI output is critical. This requires robust design principles and architectural changes to the entire AI pipeline so as to engrain trust as one of the core dimensions for enhancing transparency and explainability of AI decisions. The core challenge in such models is how to enable AI-based cognitive computing that balances robustness with trustworthy explainability. Essentially, the former requires not only the curation of data as it pertains to a harmonization pipeline so as to be able to construct a rather generic data model based on which various-transactional AI agents can operate, but also the enrichment of such data processing through a "trust plane" that can characterize/label the trustworthiness of the data source. Accounting for the heterogeneity of data sources (be it devices or entire service-graph-chains), ensuring strict integrity and assurance mechanisms is not straightforward. What is needed is a custom Trusted Computing Base that can be agnostic to the underlying Root-of-Trust. Be it a TEE or another HW-based RoT, such a TCB will be able to expose abstract interfaces for quantifying the trustworthiness level of the data sources functionality. Such a TCB will enable the transofrmation of AI Agentic Constellation into a Trusted Agentic AI Constellation. This is not only for the establishment of secure and authenticated channels but also for the verifiable and designated state management of all AI agents; i.e., the secure sharing and synchronization of models, parameters and other configuration of training data models amongst this federation of AI agents. All in all, safety assurance and dynamic trust assessment frameworks are crucial to validate AI models against rigorous safety standards, addressing potential failures and ensuring safe operations in critical environments. Compundng this issue, research for the creation of new approaches and standards is required towards the creation of intelligent networks and agentic web with particular emphasis on policy-based authorization, delegation of entitlements, and trustworthiness assessment of data paths in intelligent networks.

Required/Preferred Candidate Skills and Competencies:

Successful candidates will demonstrate academic excellence and outstanding research potential. Applicants should have a two-year Master’s degree or a similar degree with an academic level equivalent to an MSc degree. A good background in the theory and practice of adversarial machine learning is essential coupled with knowledge in trusted computing, intelligent networks towards the establishment of Trustworthy AI management. Exposure also to applied crypto and digital identity for transforming AI-based outputs into security or safety policies and/or enforcing security controls is optional but beneficial. Good implementation skills and practical experience are also desirable. Furthermore, good command of the English language is essential.

The intellectual property of the research results that will derive from the activities carried out by the doctoral student is owned by the Company.

Contact: Bruno Crispo (UniTrento),Thanassis Giannetsos (Ubitech)

IMG > Adobe Stock, ©ipopba