2019 - 35th cycle
Materials Science and Engineering (area A)
- Reference persons: I. Cristofolini, A. Molinari
Title: Development of a model of the anisotropic dimensional change on sintering. 1)
The aim of the project is the development of a model for the prediction of dimensional change on sintering of prior cold compacted green parts. Dimensional change (either shrinkage or swelling) is affected by the material, the compaction strategy, the green density and the parameters of the sintering process. It is anisotropic, and its anisotropy is also affected by the geometry of the parts. A model has been developed in previous works, which can predict dimensional changes of real parts with quite a good approximation, corresponding to half of the usual ISO IT tolerance class in Powder Metallurgy. The project aims at improving the accuracy of the model. To this purpose, a wide experimental campaign will be carried out in cooperation with some of the biggest PM industries in Europe, to generate a comprehensive database based on the knowledge of the mechanisms responsible for dimensional change and its anisotropy. In parallel, the model previously developed will be improved to increase its robustness, analysing the physical phenomena occurring during the whole of the PM process (including cold compaction), as affected by the geometry and the green density of the parts. The model will be verified by designing some real parts.
- Reference persons: A. Dorigato, L. Fambri
Title: Multifunctional polymeric foams: preparation, characterization and environmental aspects. 1)
Due to their low density and elevated mechanical properties, structural polymeric foams can find application in a wide variety of sectors, from the construction to the technical packaging field. In some cases, the porous structure of these materials can be generated through the addition of ceramic hollow microspheres (i.e. syntactic foams), while in other cases elevated stiffness and strength values can be obtained through the preparation of sandwich structures, interposing these foams between metallic and/or composite laminates. In the greatest part of the works present in literature, the attention was mainly focused on the evaluation of the mechanical behaviour of the foams, while less attention has been devoted to the possibility of developing porous polymeric structures combining high mechanical performances and functional properties (i.e. flame retardancy, high electrical conductivity, self sensing/healing behaviour, thermal energy storage capability). Therefore, the aim of this project is that to develop novel multifunctional polymer foams for structural applications, combining elevated stiffness/strength to functional properties. In this work, different polymeric systems (i.e. thermoplastics, thermosettings) and different foam structures (syntactic foams, sandwich composites) will be considered. Due to the increasing environmental concerns about the use of plastic materials and relative foams, also matrices coming from recycled waste streams and/or synthesized from renewable resources will be analyzed. A comprehensive characterization of the physical behaviour of the resulting foams will be performed, in order to highlight the multifunctional potential of the produced materials. The analysis of the environmental performance of these foams will be assessed through Life Cycle Assessment (LCA) technique.
- Reference persons: S. Rossi, V. Fontanari
Title: Innovative composite enamel coatings with improved mechanical properties. 1)
The vitreous enamel coatings show excellent corrosion protection properties for steel and aluminum substrate. In addition, this deposits present good chemical resistance, resistance at high temperature and interesting aesthetical properties. Due to their vitreous nature, these coatings shows a brittle behavior and limited mechanical properties, which limits the application of this kind of layer. The aim of the research is to improve the mechanical behavior with the modification of chemical composition of frits and of starting materials, the introduction of hard particles and reinforce as graphene flakes to improve the abrasion resistance and fracture toughness. The influence of the application methods and of the firing temperature and will be considered. Other possible solutions for proprerties improvement, identified by consulting the literature, could be taken into consideration. The corrosion protection properties and other useful properties typical of this kind of coatings should be maintained or improved.
- Reference persons: S. Dirè, M. Fontana, A. Pegoretti
Title: Development of smart polymer-ceramic composites with piezoelectric behaviour. 1)
Piezoelectric materials are employed in advanced applications, such as high-energy-density capacitors, transducers, actuators, sensors. Among smart materials, composites characterized by piezoelectric behaviour appear the most promising candidates to be integrated in devices for energy harvesting applications. The aim of the research is to produce flexible smart piezoelectric composites by properly embedding ferroelectric ceramic fillers (even if less performing respect to PZT, PMN and similar phases lead-free perovskites will be preferentially considered for more environmentally safe applications), in suitable polymer matrices (such as PDMS, PVDF,…) characterized by high breakdown field. Surface modification of the ceramic particles as well as control of microstructure will allow to optimize energy harvesting and storage capabilities of the resultant composites, while retaining processability and deformability typical of polymers. The possibility to combine the ceramic particles with other fillers, acting as dispersant and reinforcing agent or providing additional functionality, will be also evaluated.
- Reference persons: A. Motta, D. Maniglio
Title: Engineering smart polymers based (bio)inks for complex 3D matrix fabrication. A1)
Hydrogels based on natural or synthetic polymers are particularly attractive for biofabrication, since they can constitute artificial extracellular matrix and generate physiological environment suitable for cell and tissue growth. In fact, hydrogels are good candidates for encapsulating cells during their fabrication process: cell-laden hydrogel can serve as building blocks or used in conjunction with a set of additive manufacturing techniques that go under the name of bioprinting. The raw materials of bioprinting process, soft biomaterials loaded with living cells, are called bioinks. They can be based on natural or synthetic polymers alone, or a combination of the two as hybrid materials. An ideal bioink should possess proper mechanical, rheological, and biological properties that should be designed, depending on the target tissues, which are essential to ensure correct functionality of the bioprinted tissues and organs.The research will focus on the formulation of blends of natural and synthetic polymers to be used as bioinks in 3D bioprinting methods. Different requirements should be considered and characteristics optimized accordingly with printing modality and target tissue type. It will consider rheological properties, extrudability through thin needles or nozzles, methods to induce the solution gelation, effect of the gelation mechanism, and mechanical properties. Chemical and physical properties will be balanced with biomaterial ink and the addition of functional chemical groups will be consider to enhance bioink performances. Biocompatibility of cells with direct contact and encapsulation will be evaluated while analysing viability, metabolism and proliferation, degradability or stability of the final gel. The PhD research will focus on the formulation, chemical functionalization, physical/biological characterization and processability of hydrogels to be used as bioinks in 3D bioprinting methods.
- Reference persons: C. Menapace, S. Gialanella, G. Straffelini
Title: Recycling vehicular brake pads. A2)
The principles and guidelines of sustainable production and circular economy are essential aspects of the industrial product development. In this regard, the complete assessment of the full life cycle of any product is a fundamental prerequisite, particularly for mass products, like vehicular brake pads. The present research, starting from the in depth analysis of the present approaches, aims at designing a recycling process of exhaust vehicular brake pads. The feasibility of the process will be assessed using LCA and LCCA. The experimental part will be based on lab size equipment, although an important output of the project is the scaling up strategy.
- Reference persons: G. Straffelini, S. Gialanella, G. Perricone
Title: Wear performance maps for friction brake materials. A3)
Nowadays, for the design of new braking systems, an integrated approach able to consider together the performance parameters along with particulate matter emissions is needed. In order to achieve this goal, the methodology based on the so-called performance maps appears quite promising. The present research starts from a deep theoretical analysis of the performance maps in brake designing, and aims at obtaining specific experimental maps for different disc-pad systems – studied at the material level - to be used for predictive purposes of the brake performances. The experimental materials will be selected according to their friction and wear behavior and their environmental impact. The wear tests will be carried out using a subscale brake dynamometer using specifically developed braking cycles. The relevant performance maps will be integrated with the characterization and modelling of the acting wear mechanisms, in order to obtain robust design relationships. Eventually, the obtained procedure will be validated with car tests.
- Reference person: V.M. Sglavo
Title: Development of innovative SOFC by colloidal processes and co-sintering to be used with biofuels. A4)
The project, included in a wider national program on SOFC to be used with biofuels, aims to realise innovative solid oxide button cells by colloidal processes (tape casting and screen printing) and co-sintering with compositions suitable to be used with biofuels. Fundamental tasks are represented by the selection of proper powder compositions for anode and the electrolyte, the preparation of proper suspensions to be used by tape casting, the assembling of the multilayer and its co-sintering. Innovative consolidation techniques like flash sintering and cold sintering will be tested in addition to traditional sintering at high temperature. The produced button cells will be characterised from physical/chemical and mechanical point of view and final electrochemical tests will be carried out.
- Reference person: G. D. Sorarù
Title: Nanostructured Porous Ceramics for Environmental and Energy Applications. A5)
The research aims at developing ceramic components with hierarchical porosity and optimized pore architecture, and testing them in applications in the environmental and energy fields. Specifically, preceramic polymers will be used for producing aerogels and foams from the replica of polymeric structures. Among others, components with porosity ranging from nano- to milli-meters will be prepared in the form of: 3D printed scaffolds, foams and aerogels. Depending on the type of application, macro/mesoporous supports will be functionalized by grafting and/or coating with zeolites, mesoporous silica, MOFs, or chemical compounds (e.g. containing amine groups). Addition of low amounts of noble and/or transition metals, depending on the catalytic reaction, will introduce active centers with suitable properties. Accordingly, components with hierarchical porosity (in the micro-, meso- and macro-size range) and tailored functionality will be obtained. These materials, may be tested for: water purification and removal of recalcitrant organic pollutants through adsorption and photo-degradation mechanisms, thermal energy storage, CO2 capture, catalytic reactions for environmental and energetic applications (Selective Catalytic Reduction of NOx at low temperature and production of synthetic natural gas through Sabatier reaction (methanation)).
Mechatronic and Mechanical Systems (area B)
- Reference persons: L. Zaccarian
Title: Hybrid Nonlinear Control Design with Applications. 1)
Historically, nonlinear control design with stability and performance guarantees is mostly based on Lyapunov tools and limited to the study of continuous-time or discrete-time dynamics. Recent powerful frameworks for representing hybrid behaviors (continuous+discrete) allowed extending these Lyapunov approaches to a broader and richer class of systems including dynamical control with resets, impacting systems, sampled-data control and estimation, controllers involving logical modes, control of power electronics, and many more challenging technological fields including non-smooth synchronization in oscillating systems, hybrid limit cycles, hybrid solutions for obstacle avoidance and so on. These application areas provide interesting challenges that can be addressed using recent theoretical results in the field of Lyapunov-based hybrid control. The goal of this PhD subject is to address some of these challenges by possibly developing novel theoretical tools that are capable of filling the gap between the existing theories and their suitability for the above mentioned applications. (for more info email to: luca.zaccarian [at] unitn.it).
- Reference persons: M. Da Lio, M. De Cecco, F. Biral
Title: Human-robot interactions based on mirroring mechanisms: Human-codriver interactions in a driving simulator and neural architecture to support perception-action paradigm. B1)
The research is a continuation of the H2020 dreams4Cars project http://www.dreams4cars.eu/en (which developed the agent able to autonomously drive and the one to interact with). The research involves the integration of this agent in a driving simulator (almost completed activity) and carrying out of a series of experiments to study the forms of collaborative driving that can be achieved when the agent is able to interpret and understand the intentions of humans and vice versa. Study of autonomous driving paradigms according to SAE automation levels and alternative paradigm study, such as the rider-horse metaphor (H-metaphor) is an objective of the research. A substantial aspect of the research will concern the extension and application of solutions based on neural networks for perception and control of robotic vehicles also in the field of autonomous industrial vehicles, such as for example variational autoencoders for perception and data fusion of different sensory sources and neural networks for the optimal collaborative control of vehicles (but not limited to).
- Reference person: A. Del Prete
Title: Exploiting supervised learning to warm-start a model predictive controller for legged robots. B3)
The goal of this research activity, to be carried out within the EU Horizon 2020 MEMMO project (http://www.memmo-project.eu/), is to develop methods to generate complex movements on complex robotic systems, such as humanoid robots carrying out semi-autonomous construction and maintenance tasks, and rehabilitation exoskeletons sustaining the locomotion of paraplegic patients. The motion generation problem exhibits a combinatorial complexity (known as the curse of dimensionality) that is not compatible with the available computational resources. To overcome this issue we suggest to store a database of the movements that the robot should generate in any possible situation. By compressing this database, we would get a memory of the motion capabilities of the robot, which we could use during the motion execution to speed up the generation of new movements. Potential subjects to be investigated by the student include:
- using the memory of motion to warm-start a model predictive controller
- implementing such control scheme on a real robot
- developing sensor-based feedback controllers using learned sensor models
- Reference person: F. Biral
Title: Artificial drivers to plan trajectories and control racing cars at the limits. B9)
The research aims at developing an artificial driver model with hierarchical architecture that is able to drive a racing car at the limits both with free road and in the presence of other opponents. The artificial driver has to be able to plan and execute maneuvers that mimic those produced by professional racing drivers. Additional characteristics include the capability to change driving style, learn the vehicle dynamics of the car it is driving and chose the best trajectory in presence of the opponents. To achieve the research goals the state of art of optimal control and graph based planning algorithms will be used but also combined most recent neural network based approaches. The performance of the artificial driver performance will be evaluated both in simulation within an accurate driving simulation environment as well as with experimental vehicles available in the laboratory including the FSAE autonomous vehicle under development. The research outcome is intended not only for racing vehicles but also to understand how to control autonomously a vehicle during emergency maneuvers.
Electronic Systems and Integrated Microelectronic Systems (area B)
- Reference person: D. Brunelli
Title: Intelligent electronic systems for Internet of Things. 1)
Internet of Things (IoT) and smart sensors will disrupt the way to conceive manufacturing and many applications. In the near future, many sectors from Smart Cities to environmental monitoring, from Smart Grids to medical applications will use the data fusion from hundreds of smart devices. New technologies, interfaces, scalability and flexibility, energy neutrality, and power management of sensors and cloud of sensors are some of the new challenges in research. The proposed Ph.D. activity aims at designing, developing, and validating new and smart embedded architectures and solutions for the Internet of Things, with the ultra-low power design as the main driver.
- Reference person: G.-F. Dalla Betta
Title: Micromachined silicon radiation detectors. 1)
This activity aims at developing novel radiation sensors based on bulk micromachining of silicon by means of Deep Reactive Ion Etching (DRIE) and/or wet etching (e.g., by TMAH). These types of sensors exploit the 3rd dimension within the silicon substrate to offer several interesting features, among them enhanced radiation tolerance, high speed, low power consumption, etc. Examples of activities in this field within INFN and EU funded projects are sensors with three-dimensional electrodes and/or active edges, for High Energy Physics and X-ray imaging applications, and hybrid detectors of thermal neutrons based on perforated silicon sensors coupled with converter materials. The PhD research activity will deal with one or more of these projects and will be focused on the design, TCAD simulation, and experimental characterization of prototypes, both in laboratory and in beam tests also in collaboration with Italian and foreign research partner institutions.
- Reference person: D. Fontanelli
Title: Cooperative social localization. 1)
This research topic is focused on development of distributed localisation algorithms for social groups of human beings walking inside a semi-structured environment. The algorithm will rely on social models of interactions for human beings, which have to be designed contextually. The objective of the research on localisation focuses on determining the position and orientation of a mobile platform that is pushed by a human beings. The idea is to follow a multi-sensory approach of extrinsic and intrinsic platform localisation algorithms that is able to exploit - whenever available - existing infrastructure deployed in the environment but also maintains some kind of survival functionality in case external infrastructure is mainly missing. In particular, collaborative localisation algorithms will be based on multiple sources of information from nearby platforms via an approach we call “social-localisation”. The social-localisation asks for models of the human behaviours in shared spaces. Two different models will be analysed: competitive and cooperative. In the competition for shared spaces, a set of mathematical models describing motion patterns that arise when a groups of people compete for shared spaces. Instead, in the cooperation between users, we are searching for algorithms that are descriptive enough to potentiality estimate the coordinated behaviour among the users generated by the social activities.
- Reference person: D. Fontanelli
Title: Machine learning techniques for measurement. 1)
In this research thread the focus is on the novel application of machine learning techniques to classical estimation problems. In particular, the possible benefits coming from the close interactions between classic and AI based approaches will be analysed in details, identifying the limits of the applicability of both. The objective of the research is to understand to what extent the novel AI-techniques can be efficiently applied to classic estimation problems, ranging from smart grid applications to sensor fusion for robots to estimation of mechanical system dynamics.
- Reference persons: D. Macii, D. Petri
Title: Instrumentation and measurements for smart grids. 1)
This research area covers the development of novel measurement systems and techniques for the smart grid. Possible resarch topics include (but are not limited to): digital signal processing algorithms for Power Quality monitoring and Phasor Measurement Units (PMUs), time synchronization of networked distributed systems, algorithms for distribution systems state estimation (DSSE), power systems modeling, optimal placement strategies of measurement instruments on the grid, real-time monitoring systems based on wired and wireless technologies.
- Reference person: D. Petri
Title: Fundamentals of measurement in the context of big-data. 1)
In recent years, acquisition of empirical data is becoming more and more ubiquitous thanks to the huge sensing, storage, computation, and transmission capabilities offered at very low cost by current Information Technology systems. A main driver of this phenomenon, known as big data, is the relevance attributed to the role of data in supporting decision making. The purpose of this research is to develop novel models and methods to identify possible sources of uncertainty that limits the quantity of information carried by data, so enabling both the estimation and the improvement of the probability of correct decisions.
- Reference person: L. Pancheri
Title: Monolithic pixel sensor for radiation imaging. 1)
The goal of this activity is the design and characterization of CMOS pixel sensors based on fully depleted silicon, tailored for particle detection, X-ray and Near Infrared imaging. Fully depleted CMOS sensors have gained the attention of the scientific community because they can provide large detector areas with high efficiency and a reduced power consumption if compared hybrid sensors. The proposed activity will be focused on the development and optimization of the in-pixel sensing device and will be part of a larger project aimed at implementing large-area array sensors for medical, astrophysics and high-energy physics applications in collaboration with research partner institutions. Devices and test structures will be simulated using TCAD tools, and their layout will be designed using commercial CAD software. The devices will be fabricated in a customized deep submicrometer CMOS process and validated through extensive experimental characterization.
- Reference person: D. Brunelli
Title: Electronic Systems for autonomous vehicles and ambient intelligence. B2)
Autonomous vehicles represent a potentially disruptive yet beneficial change for many sectors of our society, from Smart Cities to environmental monitoring to transportation. Moreover, the synergy with the environment enhanced with Artificial Intelligence will be the breakthrough technology in the next future. The proposed Ph.D. activity aims at designing, developing, smart electronic, and embedded solutions for this area.
- Reference person: D. Fontanelli
Title: AI-based estimation and control algorithms for robotics. B4)
In many robotic applications, a robot is required to operate within unknown or partially known environments. In such operational scenarios standard estimation approaches fall short of the application requirements, as they do not provide essential elements of information to take "informed" control decisions. As an example, in human-robot interactions, estimating the pose of the human hands is not necessarily sufficient to decide what the robot is supposed to do. On the other hand, relying solely on AI techniques (e.g., deep learning, or supervised learning) any formal proof of correctness for the robot's behaviour cannot be assessed. The hypotheses underpinning the present research project is that a reliable approach for a robot to deal with uncertain, unstructured and noisy environment is at the confluence between classical estimation theory and machine learning. In this respect, a number of benchmark problems concerning mobile robotics and/or human/robot interactions will be considered and the possible application of both classical and AI based estimation algorithm exposing the limitations of each explored. The expected final outcome of the project is to show how appropriate synergies between AI and classical estimation theory can significantly amplify the benefits of each.
- Reference persons: D. Brunelli, M. Bernard
Title: Electronic control systems for integrated quantum photonics. B5)
The core of the Ph.D. activity will be to develop and test a microprocessor-based system to control integrated photonic quantum circuits. The control and manipulation of the photonic system will be realized by using integrated metallic resistors (heaters). Moreover, the activity will address a control system for the fine-tuning of the Photonic Integrated Circuit (PIC), the readout of the integrated detector arrays, as well as study other possible feedback channels form the chip.
- Reference persons: L. Gasparini, L. Pancheri, A. Stefanov
Title: CMOS-SPAD camera for entangled photon detection. B6)
The goal of this activity is to develop a new concept of single photon detector for measuring the spatio-temporal correlations existing in a flux of coincident photons. Such a detector would find application in quantum science, e.g. for the detection of entangled photons in quantum imaging or the generation of quantum random numbers in quantum communication. The concept will first be validated using an existing CMOS-SPAD imager and a source of entangled photons. An optimized detector will then be designed to maximize the overall performance of the camera and used to implement a complete imaging system including hardware, firmware and software. The camera performance will be certified in a real quantum optics experiment in collaboration with the University of Bern.
- Reference person: L. Lorenzelli, G.-F. Dalla Betta
Title: Devices and systems for organoids and 3D cell culture-on-a-chip. B7)
Studying how the human organs (e.g. brain) process and store information requires a continuous improvement in the in-vitro methodologies and models for monitoring cells. Different approaches have been attempted to develop technologies for performing in vitro tests by isolating individual cells and neurons or by growing cell cultures on planar chips based on Micro Electrode Arrays (MEAs) or sensors-based microsystems (Lab on a Cell). Recently, the interest shifted towards new in-vitro methods involving three-dimensional neuronal structures and organoids. Organoids are a simplified versions of an organ produced in vitro in three dimensions that mimic realistically the microanatomy. Assessing the neural connectivity and/or cell electrophysiological paths has become thus an important research objective that requires innovative platforms for neurons’ and cells signal recording. The objective of the proposed research is to investigate new technologies allowing a better understanding of the neurons electrical interaction and cell metabolism at the three-dimensional level.
- Reference person: F. Antonelli, D. Brunelli
Title: Distributed Intelligence in the Industrial IoT Edge. B8)
This scholarship is focused on the Industrial Internet of Things (IIoT) as the application of the IoT to the industrial domain. In this context, there are interconnected sensors, equipment, processes, and people producing, transmitting, storing, analyzing and consuming tremendous volumes of data. Here, typical cloud-based approaches cannot allow for the required levels of responsiveness, flexibility, and manageability, hence novel computing paradigms are needed for this information to be processed closer to its source in a more reliable, scalable, real-time and secure way. This is in essence what Edge Computing is asked to address to truly realize the Industry 4.0 vision. One of the major challenges of this paradigm is the distribution of complex applications’ building blocks, which includes artificial intelligence-related tasks (e.g., training, feature extraction, filtering, and model inference) on the so-called cloud-to-thing continuum. The objective of this scholarship is to enable full distribution of commonly used machine learning blocks on complex and heterogeneous IT stacks of real IIoT scenarios.
Operational Research (area B)
- Reference person: M. Fedrizzi
Title: Pairwise Comparison Matrices in Decision Making. 1)
The method of pairwise comparison is widely used in studying preferences of the decision makers, especially in multi-criteria decision making (MCDM). Some of the most relevant problems in using pairwise comparison matrices are the priority derivation and the consistency evaluation. In the last decades, many approaches have been proposed to study these problems. The most popular and probably most frequently used approach refers to Saaty’s AHP, where priorities for the given alternatives are derived by means of the maximum eigenvector of a pairwise comparison matrix and consistency is evaluated by means of the so-called consistency index. Nevertheless, numerous other researchers criticized Saaty’s approach and developed both different methods to derive the weight vector for the alternatives and new consistency indices. The first part of the research will focus on studying and classifying the different methods proposed in the literature by evidencing their most relevant characteristics. The second part of the research will propose some new approaches based on the results obtained in the first part. The study will also take into account the important aspect of measurement scales and the relationship between cardinal consistency and ordinal consistency (transitivity). Numerical computations will play a crucial role in the research project, so that skills in programming, use of software and numerical methods is highly desirable, as well as a solid mathematical background.
A1) UNITN and DII project (A. Motta)
A2) UNITN and DII project (S. Gialanella)
A3) DII project “Brembo” (G. Straffelini)
A4) DII project (V.M. Sglavo)
A5) DII project (G.D. Sorarù), project financed by MIUR PRIN 2017 CUP no. E64I19001220001
B1) UNITN and EU Projects “Dreams4Car” (M. Da Lio) & “Safe Strip” (F. Biral)
B2) UNITN and DII project “UE GREENDATANET” (D. Brunelli)
B3) EU H2020 Project “MEMMO” (A. Del Prete)
B4) EU H2020 Project “ACANTO” (D. Fontanelli)
B5) Q@TN 1
B6) Q@TN 2
B7) FBK 1
B8) FBK 2
B9) DII project (F. Biral)