Robert Bosch (Slovenia)

This work presents Neural Equivariant Interatomic Potentials (NequIP), an E(3)-equivariant neural network approach for learning interatomic potentials from ab-initio calculations for molecular dynamics simulations. While most contemporary symmetry-aware models use invariant convolutions and only act on scalars, NequIP employs E(3)-equivariant convolutions for interactions of geometric tensors, resulting in a more information-rich and faithful representation of atomic environments. The method achieves state-of-the-art accuracy on a challenging and diverse set of molecules and materials while exhibiting remarkable data efficiency. NequIP outperforms existing models with up to three orders of magnitude fewer training data, challenging the widely held belief that deep neural networks require massive training sets. The high data efficiency of the method allows for the construction of accurate potentials using high-order quantum chemical level of theory as reference and enables high-fidelity molecular dynamics simulations over long time scales.

A summary of the most important technological applications employing reduced graphene oxide.

A simultaneously accurate and computationally efficient parametrization of the potential energy surface of molecules and materials is a long-standing goal in the natural sciences. While atom-centered message passing neural networks (MPNNs) have shown remarkable accuracy, their information propagation has limited the accessible length-scales. Local methods, conversely, scale to large simulations but have suffered from inferior accuracy. This work introduces Allegro, a strictly local equivariant deep neural network interatomic potential architecture that simultaneously exhibits excellent accuracy and scalability. Allegro represents a many-body potential using iterated tensor products of learned equivariant representations without atom-centered message passing. Allegro obtains improvements over state-of-the-art methods on QM9 and revMD17. A single tensor product layer outperforms existing deep MPNNs and transformers on QM9. Furthermore, Allegro displays remarkable generalization to out-of-distribution data. Molecular simulations using Allegro recover structural and kinetic properties of an amorphous electrolyte in excellent agreement with ab-initio simulations. Finally, we demonstrate parallelization with a simulation of 100 million atoms.

<div class="htmlview paragraph">Although the total number of car occupants involved in accidents in Germany has not significantly reduced during the past 10 years, the number of fatalities has steadily decreased. Most of the severe accidents result from a loss of control of the car. The problem of the driver losing control of his car will be explained. This problem is then used to formulate the goal for the vehicle dynamics control system ESP (<u>E</u>lectronic <u>S</u>tability <u>P</u>rogram, also known as VDC). The approach chosen to reach this goal will then be shown. It will be shown that the vehicle slip angle is a crucial indicator for the maneuverability of the automobile. Since the complete vehicle state is not readily available, estimation algorithms are used to supply the control algorithms with sufficient information. With the automatic control of the slip angle the required yaw moment can be generated by individual wheel slip control. By using two examples it will be shown, that ESP can significantly improve vehicle handling in extreme maneuvers by automatically controlling the brakes and the engine.</div>

Defects are important features in two-dimensional (2D) materials that have a strong influence on their chemical and physical properties. Through the enhanced chemical reactivity at defect sites (point defects, line defects, etc.), one can selectively functionalize 2D materials via chemical reactions and thereby tune their physical properties. We demonstrate the selective atomic layer deposition of LiF on defect sites of h-BN prepared by chemical vapor deposition. The LiF deposits primarily on the line and point defects of h-BN, thereby creating seams that hold the h-BN crystallites together. The chemically and mechanically stable hybrid LiF/h-BN film successfully suppresses lithium dendrite formation during both the initial electrochemical deposition onto a copper foil and the subsequent cycling. The protected lithium electrodes exhibit good cycling behavior with more than 300 cycles at relatively high coulombic efficiency (>95%) in an additive-free carbonate electrolyte.

The ever-growing availability of computing power and the sustained development of advanced computational methods have contributed much to recent scientific progress. These developments present new challenges driven by the sheer amount of calculations and data to manage. Next-generation exascale supercomputers will harden these challenges, such that automated and scalable solutions become crucial. In recent years, we have been developing AiiDA (aiida.net), a robust open-source high-throughput infrastructure addressing the challenges arising from the needs of automated workflow management and data provenance recording. Here, we introduce developments and capabilities required to reach sustained performance, with AiiDA supporting throughputs of tens of thousands processes/hour, while automatically preserving and storing the full data provenance in a relational database making it queryable and traversable, thus enabling high-performance data analytics. AiiDA's workflow language provides advanced automation, error handling features and a flexible plugin model to allow interfacing with external simulation software. The associated plugin registry enables seamless sharing of extensions, empowering a vibrant user community dedicated to making simulations more robust, user-friendly and reproducible.

ABSTRACT We investigate optimal intertemporal asset allocation and location decisions for investors making taxable and tax‐deferred investments. We show a strong preference for holding taxable bonds in the tax‐deferred account and equity in the taxable account, reflecting the higher tax burden on taxable bonds relative to equity. For most investors, the optimal asset location policy is robust to the introduction of tax‐exempt bonds and liquidity shocks. Numerical results illustrate optimal portfolio decisions as a function of age and tax‐deferred wealth. Interestingly, the proportion of total wealth allocated to equity is inversely related to the fraction of total wealth in tax‐deferred accounts.

Monitoring electricity consumption in the home is an important way to help reduce energy usage and Non-Intrusive Load Monitoring (NILM) techniques are a promising approach to obtain estimates of the electrical power consumption of individual appliances from aggregate measurements of voltage and/or current in the distribution system. In this paper, we discuss event detection algorithms used in the NILM literature and propose new metrics for evaluating them. In particular, we introduce metrics that incorporate information contained in the power signal instead of strict detection rates. We show that this information is important for NILM applications with the goal of improving appliance energy disaggregation. Our work was carried out on a publicly-available week-long dataset of real residential power usage.

Most companies view acquisitions and mergers as onetime events managed with heroic effort--anxiety-producing experiences that often result in lost jobs, restructured responsibilities, derailed careers, and diminished power. Little wonder, then, that most managers think about how to get them over with--not how to do them better. But even as the number of mergers and acquisitions rises in the United States, studies show the performance of the resulting companies falls below industry averages more often than not. To improve these statistics, executives need to view acquisition integration as a manageable process, not a unique event. One company that has done exactly that is GE Capital Services, which has assimilated more than 100 acquisitions in the past five years alone and, in the process, has developed a formal model for melding new acquisitions into the corporate fold. Drawing on their experiences working with the company to develop the model, consultants Ron Ashkenas and Suzanne Francis, together with GE Capital's Lawrence DeMonaco, offer four lessons from the company's successful run. First, begin the integration process before the deal is signed. Second, dedicate a full-time individual to managing the integration process. Third, implement any necessary restructuring sooner rather than later. And fourth, integrate not only the business operations but also the corporate cultures. These guidelines won't erase all of the discomfort that accompanies many mergers, but they can make the process more transparent and predictable for those involved.

The evolution of the Internet into the information highway offers many future possibilities and potential problems in the delivery of counseling services. Features of the current Internet are briefly reviewed, and the results of an analysis of counseling applications on the Internet are presented and discussed. Current experience with computer networks, counseling applications, and the likely features of the future information highway provides a foundation for exploring the potential benefits and limitations of using this technology in counseling.

Atomic layer deposition enables stable cycling of Li metal in a three-dimensional lithium host.

Molecular dynamics simulation is an important tool in computational materials science and chemistry, and in the past decade it has been revolutionized by machine learning. This rapid progress in machine learning interatomic potentials has produced a number of new architectures in just the past few years. Particularly notable among these are the atomic cluster expansion, which unified many of the earlier ideas around atom-density-based descriptors, and Neural Equivariant Interatomic Potentials (NequIP), a message-passing neural network with equivariant features that exhibited state-of-the-art accuracy at the time. Here we construct a mathematical framework that unifies these models: atomic cluster expansion is extended and recast as one layer of a multi-layer architecture, while the linearized version of NequIP is understood as a particular sparsification of a much larger polynomial model. Our framework also provides a practical tool for systematically probing different choices in this unified design space. An ablation study of NequIP, via a set of experiments looking at in- and out-of-domain accuracy and smooth extrapolation very far from the training data, sheds some light on which design choices are critical to achieving high accuracy. A much-simplified version of NequIP, which we call BOTnet (for body-ordered tensor network), has an interpretable architecture and maintains its accuracy on benchmark datasets.

A critical enabling technology for electrified vehicles and renewable energy resources is battery energy storage. Advanced battery systems represent a promising technology for these applications, however their dynamics are governed by relatively complex electrochemical phenomena whose parameters degrade over time and vary across manufacturer. Moreover, limited sensing and actuation exists to monitor and control the internal state of these systems. As such, battery management systems require advanced identification, estimation, and control algorithms. In this paper we examine a new battery state-of-charge (SOC) estimation algorithm based upon the backstepping method for partial differential equations (PDEs). The estimator is synthesized from the so-called single particle model (SPM). Our development enables us to rigorously analyze observability and stability properties of the estimator design. In a companion paper we examine state-of-health (SOH) estimation, framed as a parameter identification problem for parabolic PDEs and nonlinearly parameterized output functions.

Social media language contains huge amount and wide variety of nonstandard tokens, cre-ated both intentionally and unintentionally by the users. It is of crucial importance to nor-malize the noisy nonstandard tokens before applying other NLP techniques. A major challenge facing this task is the system cov-erage, i.e., for any user-created nonstandard term, the system should be able to restore the correct word within its top n output candi-dates. In this paper, we propose a cognitively-driven normalization system that integrates different human perspectives in normalizing the nonstandard tokens, including the en-hanced letter transformation, visual priming, and string/phonetic similarity. The system was evaluated on both word- and message-level using four SMS and Twitter data sets. Results show that our system achieves over 90 % word-coverage across all data sets (a 10 % absolute increase compared to state-of-the-art); the broad word-coverage can also successfully translate into message-level per-formance gain, yielding 6 % absolute increase compared to the best prior approach. 1

Common statistical prediction models often require and assume stationarity in the data. However, in many practical applications, changes in the relationship of the response and predictor variables are regularly observed over time, resulting in the deterioration of the predictive performance of these models. This paper presents Linear Four Rates (LFR), a framework for detecting these concept drifts and subsequently identifying the data points that belong to the new concept (for relearning the model). Unlike conventional concept drift detection approaches, LFR can be applied to both batch and stream data; is not limited by the distribution properties of the response variable (e.g., datasets with imbalanced labels); is independent of the underlying statistical-model; and uses user-specified parameters that are intuitively comprehensible. The performance of LFR is compared to benchmark approaches using both simulated and commonly used public datasets that span the gamut of concept drift types. The results show LFR significantly outperforms benchmark approaches in terms of recall, accuracy and delay in detection of concept drifts across datasets.

As billions of personal data being shared through social media and network, the data privacy and security have drawn an increasing attention. Several attempts have been made to alleviate the leakage of identity information from face photos, with the aid of, e.g., image obfuscation techniques. However, most of the present results are either perceptually unsatisfactory or ineffective against face recognition systems. Our goal in this paper is to develop a technique that can encrypt the personal photos such that they can protect users from unauthorized face recognition systems but remain visually identical to the original version for human beings. To achieve this, we propose a targeted identity-protection iterative method (TIP-IM) to generate adversarial identity masks which can be overlaid on facial images, such that the original identities can be concealed without sacrificing the visual quality. Extensive experiments demonstrate that TIP-IM provides 95%+ protection success rate against various state-of-the-art face recognition models under practical test scenarios. Besides, we also show the practical and effective applicability of our method on a commercial API service.

This study compared performance on the Functional Independence Measure for Children (WeeFIM), the Battelle Developmental Inventory Screening Test (BDIST), and the Vineland Adaptive Behavior Scales (VABS) in children with developmental disabilities. The three instruments were administered to 205 children with identified disabilities. All 205 children were tested using the WeeFIM instrument. The BDIST was administered to 101 children and the VABS to the remaining 104 children. Administration was counterbalanced and randomized across all three instruments. A proportional sampling plan was used to select the 205 children, who ranged in age from 11 to 87 months. A variety of medical diagnoses and levels of severity of motor, cognitive, and communication impairments were systematically included in the sample. Correlations (r) among subscales for all three instruments ranged from 0.42 to 0.92. Correlations for total scores ranged from 0.72 to 0.94. Analyses of potential moderator variables found no significant relation between age and severity of disability (r=0.05) or between socioeconomic status (SES) and severity of disability (r=0.21). Correlations with age were strongest for those subscale scores involving gross and fine motor skills. Correlations with SES and subscale scores ranged from 0.03 to 0.18. The three instruments provide important information regarding childhood performance in motor, self-care, communicative, cognitive, and social skills. The WeeFIM instrument requires less administration time and provides information directly relevant to evaluating functional outcomes for children with disabilities and their families.

This paper presents a pixel pitch-matched readout chip for 3-D photoacoustic (PA) imaging, featuring a dedicated signal conditioning and delta-sigma modulation integrated within a pixel area of 250 µm by 250 µm. The proof-of-concept receiver was implemented in an STMicroelectronics's 28-nm Fully Depleted Silicon On Insulator technology, and interfaces to a 4 × 4 subarray of capacitive micromachined ultrasound transducers (CMUTs). The front-end signal conditioning in each pixel employs a coarse/fine gain tuning architecture to fulfill the 90-dB dynamic range requirement of the application. The employed delta-sigma beamforming architecture obviates the need for area-consuming Nyquist ADCs and thereby enables an efficient in-pixel A/D conversion. The per-pixel switched-capacitor ΔΣ modulator leverages slewing-dominated and area-optimized inverter-based amplifiers. It occupies only 1/4th of the pixel, and its area compares favorably with state-of-the-art designs that offer the same SNR and bandwidth. The modulator's measured peak signal-to-noise-and-distortion ratio is 59.9 dB for a 10-MHz input bandwidth, and it consumes 6.65 mW from a 1-V supply. The overall subarray beamforming approach improves the area per channel by 7.4 times and the single-channel SNR by 8 dB compared to prior art with similar delay resolution and power dissipation. The functionality of the designed chip was evaluated within a PA imaging experiment, employing a flip-chip bonded 2-D CMUT array.

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.

Existing neural network-based autonomous systems are shown to be vulnerable against adversarial attacks, therefore sophisticated evaluation of their robustness is of great importance. However, evaluating the robustness under the worst-case scenarios based on known attacks is not comprehensive, not to mention that some of them even rarely occur in the real world. Also, the distribution of safety-critical data is usually multimodal, while most traditional attacks and evaluation methods focus on a single modality. To solve the above challenges, we propose a flow-based multimodal safety-critical scenario generator for evaluating decision-making algorithms. The proposed generative model is optimized with weighted likelihood maximization and a gradient-based sampling procedure is integrated to improve the sampling efficiency. The safety-critical scenarios are generated by efficiently querying the task algorithms and a simulator. Experiments on a self-driving task demonstrate our advantages in terms of testing efficiency and multimodal modeling capability. We evaluate six Reinforcement Learning algorithms with our generated traffic scenarios and provide empirical conclusions about their robustness.