Accenture (United States)

<title>Abstract</title> Lattice viscosity quantifies the irreversible relaxation of momentum and stress mediated by phonon dynamics and provides a sensitive probe of anharmonicity and dissipation in crystalline solids. In ferroelectric and quantum-paraelectric materials, soft polar phonon modes dominate this response and are strongly coupled to external electric fields. Here we develop a microscopic theory of <bold>electric-field-dependent lattice viscosity</bold> based on linear response and phonon Green’s functions, explicitly incorporating anharmonic phonon–phonon interactions and nonlinear polarization effects. We show that, at temperatures above the ferroelectric transition, electric fields generically induce <bold>quadratic renormalization of phonon linewidths and frequencies</bold> , leading to a universal field dependence of the lattice viscosity governed by a small set of effective coefficients. We validate these predictions using <bold>first-principles density functional perturbation theory (DFPT)</bold> calculations implemented in Quantum ESPRESSO for representative systems, including the quantum paraelectric SrTiO₃ and the displacive ferroelectric BaTiO₃. The computed phonon dispersions, anharmonic linewidths, and inelastic scattering cross sections confirm the predicted quadratic field scaling, demonstrate strong mode selectivity associated with soft polar modes, and establish the microscopic origin of the dominant viscosity contributions. Based on these results, we establish a <bold>predictive model for the lattice viscosity Λ(E,T)</bold> that maps first-principles phonon properties directly to the electric-field and temperature dependence of lattice viscosity. Our findings reveal electric field as an effective and reversible control parameter for lattice dissipation, providing a predictive framework for electric-field control of dissipation in functional and quantum materials. We discuss implications for ferroelectric quantum criticality, phonon-mediated superconductivity, and field-tunable dissipation in oxide heterostructures and quantum devices, positioning lattice viscosity as a controllable transport property in quantum and functional materials.

The recent popularity of virtual reality has generated an increased demand for high fidelity 3D-modeling. 3D Gaussian splatting-based methods generate novel views in open-world 3D-scenes by learning to assign Gaussian primitives to structures present in associated 2D images. Renders of novel viewpoints are then produced by “splatting” Gaussians associated with a given viewpoint onto a 2D canvas. Areas of the scene represented by a greater number of Gaussian primitives typically correspond to higher fidelity. However, due to hardware constraints, limits are imposed on the maximal number of Gaussians used to represent the scene. Therefore, to ensure optimal positional allocation of the available Gaussians, we propose using semantic segmentation to partition the Gaussian splatting optimization according to the semantic scene partition. In outdoor scenes, nearby structures that appear across multiple training images are prioritized, while regions such as the sky or low value semantic regions are deprioritized. Ultimately, allocating the Gaussians with semantic-based prioritization ensures that the 3D-scene is represented in a manner that mostly aligned with human preferences.

While Generative AI stands to be one of the fastest adopted technologies ever, studies have made evident that the usage of Large Language Models (LLMs) puts significant burden on energy grids and our environment. It may prove a hindrance to the Sustainability goals of any organization. A crucial step in any Sustainability strategy is monitoring or estimating the energy consumption of various components. While there exist multiple tools for monitoring energy consumption, there is a dearth of tools/frameworks for estimating the consumption or carbon emissions. Current drawbacks of both monitoring and estimation tools include high input data points, intrusive nature, high error margin, etc. We posit that leveraging emerging LLM benchmarks and related data points can help overcome aforementioned challenges while balancing accuracy of the emission estimations. To that extent, we discuss the challenges of current approaches and present our evolving framework, R-ICE, which estimates prompt level inference carbon emissions by leveraging existing state-of-the-art(SOTA) benchmark. This direction provides a more practical and non-intrusive way to enable emerging use-cases like dynamic LLM routing, carbon accounting, etc. Our promising validation results suggest that benchmark-based modelling holds great potential for inference emission estimation and warrants further exploration from the scientific community.

Existing navigation systems mostly consider “success” when the robot reaches within 1 m radius to a goal. This precision is insufficient for emerging applications where a robot needs to be positioned precisely relative to an object for downstream tasks, such as docking, inspection, and manipulation. To this end, we design and implement <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Aim-My-Robot</i> (<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">AMR</i>), a local navigation system that enables a robot to reach any object in its vicinity at the desired relative pose, with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">centimeter-level accuracy</i>. AMR achieves high accuracy and robustness by leveraging multi-modal sensors, precise action prediction, and is trained on large-scale photorealistic data generated in simulation. AMR shows strong sim2real transfer and can adapt to different robot kinematics and unseen objects with little to no fine-tuning.

Integrating SAP Enterprise Resource Planning (ERP) systems with Amazon Web Services (AWS) is transforming educational institutions to change the face of operational competency and ways of knowledge delivery. It will discuss the cloud-based infrastructure and how it benefits educational organizations in terms of scalability, cost optimization, and robust disaster recovery strategies. This article will assess how SAP on AWS implements automation processes in administrative work, data-driven decision-making, and personalized learning experiences. The article addresses critical security and compliance considerations indispensable for protecting sensitive educational data while maintaining regulatory adherence. It addresses best practices around implementation, from architecture planning down to performance-optimization strategy and change-management approach. To top it off, the article analyses the future scope of educational technology, which is very much concerned with the emerging themes of artificial intelligence, the Internet of Things, blockchain, and extended reality, leading to a redrawing of educational landscapes. The article provides insight into how educational institutions can use these technological advancements to enhance operational efficiency while maintaining security and compliance in an increasingly digital educational environment.

With sophisticated machine learning, deep fakes are made so real that they undermine political processes of trust by manipulating public opinion to destabilize economic stability. Persons can create false videos or audio of public figures, hence swaying elections or spreading disinformation. They can be used in defrauding, manipulating markets, and damaging reputations. Deepfake detection requires intricate algorithms to distinguish manipulation in media. Mitigation includes legal frameworks, educational initiatives, and technological solutions. Legal measures have to address misapplications with clear regulations and penalties. Academic activities are necessary to increase the media literacy of the public to identify deep fakes. However, by implementing such strategies, deep fakes could be significantly reduced in their capacity to make destabilizing influences on politics and the economy through democratic processes and create trust in digital interactions.