Search

Consiglio Nazionale delle Ricerche-CARI-CARE-ITALY’ within the CRUI CARE Agreement

This paper evaluates the yield of a memristor-based crossbar array of artificial neural networks in the presence of stuck-at-faults (SAFs). A technique based on Markov chains is used to estimate the yield in the presence of stuck-at-faults. This method provides a high degree of accuracy. Another method that is used for analysis and comparison is the Poisson distribution, which uses the sum of all repairable fault patterns. A fault repair mechanism is also considered when evaluating the yield of the memristor crossbar array. The results demonstrate that the yield could be improved with redundancies and a higher repairable stuck-at-fault ratio.

Approximate Computing (AxC) techniques allow trade-off accuracy for performance, energy, and area reduction gains. One of the applications suitable for using AxC techniques are the Spiking Neural Networks (SNNs). SNNs are the new frontier for artificial intelligence since they allow for a more reliable hardware design. Unfortunately, this design requires some area minimization strategies when the target hardware reaches the edge of computing. In this work, we first extract the computation flow of an SNN, then employ Interval Arithmetic (IA) to model the propagation of the approximation error. This enables a quick evaluation of the impact of approximation. Experimental results confirm the model’s adherence and the capability of reducing the exploration time.

Spiking Neural Networks (SNN) are an emerging type of biologically plausible and efficient Artificial Neural Network (ANN). This work presents the development of a hardware accelerator for a SNN for high-performance inference, targeting a Xilinx Artix-7 Field Programmable Gate Array (FPGA). The model used inside the neuron is the Leaky Integrate and Fire (LIF). The execution is clock-driven, meaning that the internal state of the neuron is updated at every clock cycle, even in absence of spikes. The inference capabilities of the accelerator are evaluated using the MINST dataset. The training is performed offline on a full precision model. The results show a good improvement in performance if compared with the state-of-the-art accelerators, requiring 215{\mu}s per image. The energy consumption is slightly higher than the most optimized design, with an average value of 13mJ per image. The test design consists of a single layer of four-hundred neurons and uses around 40% of the available resources on the FPGA. This makes it suitable for a time-constrained application at the edge, leaving space for other acceleration tasks on the FPGA., Comment: 6 pages, 3 figures, 4 tables

Energy production from distributed renewable power plants underwent a takeoff in last years as never before. Nevertheless, the installation of technologies based on variable energy resources and their connection on transmission power lines might cause congestions due to the transmission capacity limits. This paper describes the modelization of a HV transmission line with local renewable production and its optimal management through an Energy Hub model. Aim of the study is to identify the optimal size of the power storage, based on an electoryzer, a hydrogen storage and a fuel cell, in order to minimize the congestion risks and to maximize the exploitation of renewable energy production.

Very recent photoluminescence studies, investigating the tetragonal-to-orthorhombic phase transition of MAPbI 3 , demonstrated the presence of residual tetragonal phase far below the transition temperature, yielding spectral signatures from quantum confined tetragonal domains. We present a theoretical model of the coexistence of tetragonal and orthorhombic MAPbI 3 , based on tight binding simulations. The tight binding parameters are derived by particle swarm optimization and the band-offset between the two crystals is obtained by first-principles calculations. The impact of the tetragonal domain dimension on the optical properties of MAPbI 3 is discussed.

Both x-ray diffraction (XRD) and photo-luminescence (PL) characterizations of methylammonium lead iodide perovskite (MAPbI 3 ) reveal signatures of a coexistence of the tetragonal and orthorhombic phases over a wide range of temperatures, suggesting that the phase transition does not happen sharply at a temperature of 164 K as reported in literature. To understand the causes of this discrepancy we investigated the evolution of the phase mixture from 300 K down to 80 K and its influence on electrical device characteristics. For this we developed a numerical approach to extract effective electrical parameters like resistivity for the mixed-phase MAPbI 3 as a function of temperature and volume ratio of the two phases. The spatial distribution of the tetragonal and orthorhombic phases and their evolution with temperature has been extracted from measured XRD maps, employing a Monte Carlo algorithm to interpolate between measured temperatures. The electrical simulations have been performed by means of the drift-diffusion model.

Interconnected devices are growing very fast in today's automotive market, providing new and complex features that cover very different domains. This vast and continuous requirement for new features brings to impact areas categorized as real-time safety-critical devices, opening the possibility to add potential vulnerabilities. By analyzing the security vulnerabilities within vehicle networks, this paper aims at proposing a new generation of a secure architecture based on Controller Area Network (CAN) called TAURUM P2T. This new architecture looks at mitigating the vulnerabilities found in the current network systems of road vehicles by introducing a low-cost and efficient solution based on the introduction of a Secure CAN network able to implement a novel key provisioning strategy. The proposed architecture has been implemented, resorting to a commercial Multi-Protocol Vehicle Interface module, and the obtained results experimentally demonstrate the approach's feasibility.

Halide Perovskite photovoltaic technology can be scaled to large area modules and panels using printing processes and laser patterning. Here, we will present the progress made to scale up from small area solar cells to modules and panels up to a dimension of 0.5 sqm. Specific efforts have been devoted to developing a deposition process out of the glove box (GB) in conventional ambient air. We transfer out of the GB several coating technologies, including blade coating and slot-die. [1], [2] To do this without penalizing efficiency and stability, a specific formulation of perovskite absorber [2], [3] and doping strategies of transporting layer have been formulated. [4] These optimizations permitted to realized perovskite solar modules with an efficiency of > 17% on an active area of 43 cm2, keeping above 90% of the initial efficiency after 800 h thermal stress at 85 °C. [4] One of the critical issues scaling the cell to module size is the control of interface properties. We demonstrated that tuning of interface properties can be successfully obtained by applying two-dimensional (2D) materials, such as graphene [5], functionalized MoS2 [6], MXenes [7] as well as 2D Perovskite.

Authentication of hardware modules connected through Controller Area Networks (CAN) in modern vehicles is becoming an increasing security issue. Untrusted modules introduced on the market may alter the secure boot infrastructure of a complex vehicle, thus completely compromising its security. This paper introduces the problem and highlights a preliminary idea for reaching better protection and preventing or limiting this category of attacks.

International audience; The computing continuum's actual trend is facing a growth in terms of devices with any degree of computational capability. Those devices may or may not include a full-stack, including the Operating System layer and the Application layer, or just facing pure bare-metal solutions. In either case, the reliability of the full system stack has to be guaranteed. It is crucial to provide data regarding the impact of faults at all system stack levels and potential hardening solutions to design highly resilient systems. While most of the work usually concentrates on the application reliability, the special session aims to provide a deep comprehension of the impact on the reliability of an embedded system when faults in the hardware substrate of the system stack surface at the Operating System layer. For this reason, we will cover a comparison from an application perspective when hardware faults happen in bare metal vs. real-time OS vs. general-purpose OS. Then we will go deeper within a FreeRTOS to evaluate the contribution of all parts of the OS. Eventually, the Special Session will propose some hardening techniques at the Operating System level by exploiting the scheduling capabilities.

Nowadays, due to technology enhancement, faults are increasingly compromising all kinds of computing machines, from servers to embedded systems. Recent advances in machine learning are opening new opportunities to achieve fault detection exploiting hardware metrics inspection, thus avoiding the use of heavy software techniques or product-specific errors reporting mechanisms. This paper investigates the capability of different deep learning models trained on data collected through simulation-based fault injection to generalize over different software applications.

The mammalian cortex contains a great variety of neuronal cells. In particular, GABAergic interneurons, which play a major role in neuronal circuit function, exhibit an extraordinary diversity of cell types. In this regard, single-cell RNA-seq analysis is crucial to study cellular heterogeneity. To identify and analyze rare cell types, it is necessary to reliably label cells through known markers. In this way, all the related studies are dependent on the quality of the employed marker genes. Therefore, in this work, we investigate how a set of chosen inhibitory interneurons markers perform. The gene set consists of both immunohistochemistry-derived genes and single-cell RNA-seq taxonomy ones. We employed various human and mouse datasets of the brain cortex, consequently processed with the Monocle3 pipeline. We defined metrics based on the relations between unsupervised cluster results and the marker expression. Specifically, we calculated the specificity, the fraction of cells expressing, and some metrics derived from decision tree analysis like entropy gain and impurity reduction. The results highlighted the strong reliability of some markers but also the low quality of others. More interestingly, though, a correlation emerges between the general performances of the genes set and the experimental quality of the datasets. Therefore, the proposed method allows evaluating the quality of a dataset in relation to its reliability regarding the inhibitory interneurons cellular heterogeneity study.

Among the new generation photovoltaic (PV), Perovskite solar cell technology reached top efficiencies in a few years. Researchers should achieve stability and process/material low cost to place the technology in the market. Among the different layers that compose a Perovskite solar cell, the hole transporting layer (HTL) plays a crucial role in terms of efficiency and stability. The reference and expensive Spiro-OMeTAD HTL has high efficiency but poor stability. The main scope of this work is to optimize the process to obtain efficient, repeatable and stable devices, by using the low cost and scalable X55 HTM (hole transporting material) that is emerging in the Perovskite solar cell field. We show a champion efficiency about 16% that is higher respect to the reference Spiro-OMeTAD case (about 15%). The results are also confirmed by IPCE spectra. Finally, we demonstrate for the first time in literature the stability of the material in humidity condition.

The Building Integration Photovoltaics (BIPV) makes possible the power generation from roofs, windows or facades. The c-Si or a-Si PV cannot fully satisfy aesthetic requirements as transparency and colouring, and facing problems as poor irradiation, shading and temperatures. Among the new generation PV, Perovskite solar cell technology reached top efficiencies in a few years together with a low-cost perspective. Different strategies related to the Perovskite layer (thickness, morphology, mesh-structure, band-gap tuning) and to alternative top-electrodes can be implemented to achieve the semi-transparent feature for BIPV applications. Here the stable triple cation CsMAFA Perovskite is used for the first time in literature to realize a semi-transparent module (about 90% geometrical fill factor) with a highly transparent sputtered ITO top-electrode. The optical and electrical characterizations of the device show an average visible transmittance (AVT) equal to 10% on the module full stack and area, and an efficiency more than 8%. A failure analysis by Dark Lock-In Thermography (DLIT) and Electroluminescence (EL) imaging indicates the route to optimize the PV device. Finally, this work provides a simple method to realize the semi-transparent CsMAFA Perovskite solar module for BIPV.

In this work we present a generalized transfer matrix method to study the effect of the photon recycling on the performance of solar cells. Photon recycling increases the charge carrier concentration in solar cells, resulting in an increase of the open circuit voltage $(V_{oc})$ . The model is based on the transfer matrix method (TMM), taking into account internal sources representing the emission from radiative recombination.

The tight binding (TB) approach represents a good trade-off between accuracy and computational burden. For this reason, it is widely used for device simulations. However, a proper description of a physical system by means of TB requires an accurate parameterization of the Hamiltonian matrix elements (HME), that is usually done by fitting over suitable properties that can be measured or computed with first-principles approaches. We show that the particle swarm optimization algorithm is a powerful tool for the parameterization of the TB HME, using the density functional theory band dispersions of bulk reference materials as a target. We discuss the results obtained for bulk MAPbI 3 perovskite in its high temperature cubic phase.

Legged robots have been highlighted as promising mobile platforms for disaster response and rescue scenarios because of their rough terrain locomotion capability. In cluttered environments, small robots are desirable as they can maneuver through small gaps, narrow paths, or tunnels. However small robots have their own set of difficulties such as limited space for sensors, limited obstacle clearance, and scaled-down walking speed. In this paper, we extensively address these difficulties via effective sensor integration and exploitation of dynamic locomotion and jumping. We integrate two Intel RealSense sensors into the MIT Mini-Cheetah, a 0.3 m tall, 9 kg quadruped robot. Simple and effective filtering and evaluation algorithms are used for foothold adjustment and obstacle avoidance. We showcase the exploration of highly irregular terrain using dynamic trotting and jumping with the small-scale, fully sensorized Mini-Cheetah quadruped robot.

Acoustic echoes retrieval is a research topic that is gaining importance in many speech and audio signal processing applications such as speech enhancement, source separation, dereverberation and room geometry estimation. This work proposes a novel approach to blindly retrieve the off-grid timing of early acoustic echoes from a stereophonic recording of an unknown sound source such as speech. It builds on the recent framework of continuous dictionaries. In contrast with existing methods, the proposed approach does not rely on parameter tuning nor peak picking techniques by working directly in the parameter space of interest. The accuracy and robustness of the method are assessed on challenging simulated setups with varying noise and reverberation levels and are compared to two state-of-the-art methods.

The remarkable high power conversion efficiency (PCE) demonstrated by small area perovskite solar cells (PSCs) are still not accomplished by a clear assessment on long-term stability, while scalability on large modules still penalized a feasible production for the photovoltaic market. In this work, we propose a route to easily produce large area (active area 70 cm2) perovskite solar module with stable performance under real working conditions. The employed graphene interface engineering (GIE) strategy, together with the use of stable polymeric hole transporting layer, allowed to achieve stabilized PCE overcoming 14% with a T 80 life-time (defined by the point at which device efficiency falls to 80% of the initial value) 27 times improved for PTAA cells compared to standard spiro-OMeTAD-based devices.