Robert Bosch (Slovenia)

Silicon carbide (SiC) is an important technological material, but its high-temperature phase diagram has remained unclear due to conflicting experimental results about congruent versus incongruent melting. Here, we employ large-scale machine learning molecular dynamics (MLMD) simulations to gain insights into SiC decomposition and phase transitions. Our approach relies on a Bayesian active learning workflow to efficiently train an accurate machine learning force field on density functional theory data. Our large-scale simulations provide direct indication that melting of SiC proceeds incongruently via decomposition into silicon-rich and carbon phases at high temperature and pressure. During cooling at high pressures, carbon nanoclusters nucleate and grow within the homogeneous molten liquid. During heating, the decomposed mixture reversibly transitions back into a homogeneous SiC liquid. The full pressure-temperature phase diagram of SiC is systematically constructed using MLMD simulations, providing new understanding of the nature of phases, resolving long-standing inconsistencies from previous experiments and yielding technologically relevant implications for processing and deposition of this material.

This paper examines the relative performance of CST Studio Suite and Keysight ADS in analyzing signal integrity characteristics of an RGMIIv2.0 interface implementation. The study evaluates both simulation platforms through systematic analysis of MAC-PHY interconnect performance, focusing on transmission line behavior, impedance control, and electromagnetic coupling effects. Prelayout modeling establishes baseline channel characteristics using identical geometric parameters, while post-layout verification compares simulation accuracy against measured performance metrics. The investigation addresses S-parameter extraction, time-domain reflectometry, and eye diagram generation capabilities of each tool. Results demonstrate notable differences in solver methodologies, with CST's field-based approach providing detailed 3D electromagnetic visualization while ADS offers streamlined circuit-level analysis with faster computational performance. CST delivers superior accuracy for complex 3D structures and crosstalk analysis, whereas ADS excels in parametric optimization and integrated design workflows. This comparative evaluation provides design engineers with practical guidance for selecting appropriate simulation tools based on specific RGMII interface design requirements and computational constraints.

Ku-band circular patch antenna design and development are presented in this study, utilizing a double U-Slot technique to enhance antenna performance. The double U-Slot configuration effectively broadens the antenna's resonant bandwidth, achieving up to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$9.63 \text{GHz}(8.33-17.96 \text{GHz})$</tex>. The radiation patterns across multiple resonant frequencies are nearly identical, indicating minimal variation in resonant modes, with a peak gain of 8.1 dBi at 11.5 GHz. The antenna is further configured into a <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$4 \times 1$</tex> linear array to evaluate its performance in array applications and investigate mutual coupling effects. The measured mutual coupling between adjacent elements is approximately −25 dB, demonstrating improved isolation due to the double U-slot design. Additionally, the influence of inter-element spacing on array performance is analyzed to understand its impact on gain and radiation characteristics.

We investigate various methods of neural network optimization with the purpose of reducing runtime and maintaining performance when deployed on embedded devices. The proposed approaches follow both improvements upon the general architecture of the deep neural models, as well as possibilities of increasing hardware utilization and employing scheduling and parallelization techniques. The presented solutions significantly improve the runtime and the behaviour of neural networks when deployed on embedded targets. The optimization strategies improve the runtime by approximately 3%-4% when applied on traditional Convolutional Neural Networks. Meanwhile, the analyzed Birds-Eye View Transformer benefited from an improvement ranging from 38% up to 83% depending on the utilized hardware and the quantization range. Thus, we stress the importance of taking advantage of the available hardware resources in order to reach the best possible results in terms of execution runtime.

&lt;p&gt; Scaled-down vehicles provide a controlled and repeatable environment for testing autonomous driving algorithms, expediting development while mitigating risks associated with full-scale vehicle testing. This paper introduces a meticulously designed scaled-down electric vehicle, emphasizing vehicle state measurement, precise control, perception sensing, and electric safety. In contrast to conventional scaled platforms built on radio-controlled cars comprising actuators lacking measurements, precision, and reliability, our approach utilizes actuators with feedback, enhancing precision and torque across all speed ranges. Apart from the conventional sensors used in the existing scaled platforms, our vehicle incorporates novel sensor modules to measure wheel angular velocities, steering angles, and battery cell voltages, which can also be seamlessly integrated into these platforms. We demonstrate the vehicle's motion with manual and autonomous operation and showcase its features for validation.&lt;/p&gt;

Deep learning models are widely used for automated driving, but achieving the goal of performance in embedded Systems-on-Chip (SoCs) has a lot of challenges due to the tradeoff between accuracy and run-time. This paper addresses the task of finding a suitable neural architecture for heterogeneous Systems-on-Chip (SoCs) in order to strike a balance between accuracy and run-time efficiency. Our approach may be used to predict the inference run-time of DNN models, which is useful not only for neural architecture search, but also for the optimization of deep learning models. We propose a method for predicting the run-time of deep learning models using Graph Neural Networks, based on the Graph Convolution Network. We transform the DNN models into an acyclic graph structure, which incorporates encoded features for all nodes. Each node represents an operation of the model and all nodes are linked together. The experimental results show that our approach achieves accurate run-time inference predictions, with an overall error between 2% and 4% for ARM CPUs.

Leveraging ab initio data at scale has enabled the development of machine learning models capable of extremely accurate and fast molecular property prediction. A central paradigm of many previous works focuses on generating predictions for only a fixed set of properties. Recent lines of research instead aim to explicitly learn the electronic structure via molecular wavefunctions from which other quantum chemical properties can directly be derived. While previous methods generate predictions as a function of only the atomic configuration, in this work we present an alternate approach that directly purposes basis dependent information to predict molecular electronic structure. The backbone of our model, Orbital Mixer, uses MLP Mixer layers within a simple, intuitive, and scalable architecture and achieves competitive Hamiltonian and molecular orbital energy and coefficient prediction accuracies compared to the state-of-the-art.

&lt;div class="section abstract"&gt;&lt;div class="htmlview paragraph"&gt;It has been largely announced that automotive industry is going through a disruption moment regarding applied powertrain technologies due to the efforts to decrease CO&lt;sub&gt;2&lt;/sub&gt; and pollutant gases emission, mainly through related legislations of different countries and regions. European and Asian future legislations are going to demand some electrification introduction, whether hybrid or fully electric, but even different technologies such as fuel cells and synthetic fuels over the next few years. In Brazil, with the upswing of biofuels use, considering a well to wheel CO&lt;sub&gt;2&lt;/sub&gt; emission calculation, the usage of hydrated ethanol or ethanol mixed up with gasoline in different proportions is a great solution for a continuous and progressive automotive fleet decarbonization, in parallel or associated with electrification, in a favorable pace for the market conditions.&lt;/div&gt;&lt;div class="htmlview paragraph"&gt;This evolution, for instance, could be better planned to avoid impacts on PowerGrid and in internal combustion engines portfolio at market, since hybrid and electric vehicles which are still too costly and depend on subsidies for sale. This paper aims to show how water injection technology enables high efficiency flex fuel engines development, focused on ethanol performance, granting to Brazil a smoother path through future national fleet electrification. The energy efficiency in Mega Joules per Kilometer, improved through higher geometrical compression ratio, enhanced up to 7.3% (from 1.59 to 1.47 MJ/km) in a natural aspirated port fuel injection vehicle and up to 4.8% (from 1.85 to 1.76 MJ/km) in a turbocharged direct injection vehicle. Following this path, Flex Fuel engines can be better explored and improved if the right technology is used. Setting up the geometrical compression ratio focusing on ethanol properties could rise both ethanol and gasoline energy efficiency, as water injection can control knocking phenomena through managing combustion chamber temperature.&lt;/div&gt;&lt;/div&gt;

Automated machine learning (AutoML) represents one of the fastest-growing fields in both industry and academia. Its popularity has increased over the last few years due to the rising demand for machine learning systems usable by non-experts. Existing open-source systems fail to fully automate the machine learning tasks, being more like a black box that generates non-reusable pipelines. We propose an end-to-end system accessible to non-specialists and data scientists who wish to effortlessly create reusable and easy to understand pipelines for any regression and classification problems. The proposed solution increases the automation level by performing an advanced data cleaning followed by a complete feature engineering process without any domain knowledge about the data set. These processes are integrated into a search that generates for each model the optimum instance of preprocessed data. Bayesian optimization (BO) with a Gaussian process (GP) maximizes the performance of these models that will be ensembled using standard and probability-based stacking. We prove that the system performs better than 43.47% of practitioners on real-world data sets, being robust to any variation of tabular data. The true potential is revealed by surpassing at least one similar tool in 50% of performed benchmarks. These results were achieved for any tested data set, regardless of its dimensionality, after a maximum of 30 minutes.

The Open Databases Integration for Materials Design (OPTIMADE) consortium has designed a universal application programming interface (API) to make materials databases accessible and interoperable. We outline the first stable release of the specification, v1.0, which is already supported by many leading databases and several software packages. We illustrate the advantages of the OPTIMADE API through worked examples on each of the public materials databases that support the full API specification.