The Large Hadron Collider (LHC) near Geneva, Switzerland became famous around the world in 2012 with the detection of the Higgs boson. The observation marked a crucial confirmation of the Standard Model of particle physics, which organizes the subatomic particles into groups similar to elements in the periodic table from chemistry. The U.S. Department of Energy's (DOE) Argonne National Laboratory has made many pivotal contributions to the construction and operation of the ATLAS experimental detector at the LHC and to the analysis of signals recorded by the detector that uncover the underlying physics of particle collisions. Argonne is now playing a lead role in the high-luminosity upgrade of the ATLAS detector for operations that are planned to begin in 2027. To that end, a team of Argonne physicists and computational scientists has devised a machine learning-based algorithm that approximates how the present detector would respond to the greatly increased data expected with the upgrade.

Comau has created an innovative, in-line testing and quality control paradigm that optimizes the construction and assembly of batteries. MI.RA/Thermography is one of the newest solutions within Comau's cutting-edge vision systems family of Machine Inspection Recognition Archetypes, named MI.RA. Designed for industrial-scale battery manufacturing, MI.RA/Thermography uses thermal imaging and artificial intelligence to perform non-invasive automated assessment and control of welded joints, to ensure battery integrity and prevent waste. Its non-destructive testing methodology protects cycle times without changing the existing manufacturing layout. Battery packs are composed of a large number of individual battery cells that are structurally held and electrically connected by numerous welded joints.

The human brain has advanced over time in countering survival instincts, harnessing intellectual curiosity, and managing authoritative ordinances of nature. When humans got an idea about the dynamics of the environment, we started with our quest to replicate nature. While the human brain discovers ways to go beyond our physical capabilities, the combination of mathematics, algorithms, computational methods, and statistical models accumulated momentum after Alan Mathison Turing built a mathematical model for biological morphogenesis, and published a seminal paper on computing intelligence. Today, AI has developed from data models for problem-solving to artificial neural networks, a computational model predicated on the structure and functions of human biological neural networks. The brain, customarily perceived as an organ of the human body, should be understood as a biologically predicated form of artificial intelligence (AI).

Researchers at the University of Pittsburgh, University of Massachusetts Amherst, and Microsoft Research India have developed a system -- WattScale -- that leverages AI to pick out the least energy-efficient buildings from a city- or region-level population. In a preprint study, they used it to show that half of the buildings in a 10,000-building data set were inefficient, in large part due to poor construction. They also emit over a third of the nation's greenhouse gases, which is more than any other sector of the economy. Solving for the disparity requires identifying buildings that are the least efficient and thus have the greatest need for improvements, but approaches that rely on the age of a building or its total energy bill don't work well; greater energy usage doesn't necessarily point to inefficiencies. WattScale aims to address this with (1) a Bayesian modeling technique that captures variable distributions governing the energy usage of a building and (2) a fault analysis algorithm that makes use of these distributions to report probable causes of inefficiency.

Bayesian optimization (BO) is an efficient and flexible global optimization framework that is applicable to a very wide range of engineering applications. To leverage the capability of the classical BO, many extensions, including multi-objective, multi-fidelity, parallelization, latent-variable model, have been proposed to improve the limitation of the classical BO framework. In this work, we propose a novel multi-objective (MO) extension, called srMO-BO-3GP, to solve the MO optimization problems in a sequential setting. Three different Gaussian processes (GPs) are stacked together, where each of the GP is assigned with a different task: the first GP is used to approximate the single-objective function, the second GP is used to learn the unknown constraints, and the third GP is used to learn the uncertain Pareto frontier. At each iteration, a MO augmented Tchebycheff function converting MO to single-objective is adopted and extended with a regularized ridge term, where the regularization is introduced to smoothen the single-objective function. Finally, we couple the third GP along with the classical BO framework to promote the richness and diversity of the Pareto frontier by the exploitation and exploration acquisition function. The proposed framework is demonstrated using several numerical benchmark functions, as well as a thermomechanical finite element model for flip-chip package design optimization.

For all the advances enabled by artificial intelligence, from speech recognition to self-driving cars, AI systems consume a lot of power and can generate high volumes of climate-changing carbon emissions. A study last year found that training an off-the-shelf AI language-processing system produced 1,400 pounds of emissions – about the amount produced by flying one person roundtrip between New York and San Francisco. The full suite of experiments needed to build and train that AI language system from scratch can generate even more: up to 78,000 pounds, depending on the source of power. But there are ways to make machine learning cleaner and greener, a movement that has been called "Green AI." Some algorithms are less power-hungry than others, for example, and many training sessions can be moved to remote locations that get most of their power from renewable sources.

The free energy plays a fundamental role in descriptions of many systems in continuum physics. Notably, in multiphysics applications, it encodes thermodynamic coupling between different fields. It thereby gives rise to driving forces on the dynamics of interaction between the constituent phenomena. In mechano-chemically interacting materials systems, even consideration of only compositions, order parameters and strains can render the free energy to be reasonably high-dimensional. In proposing the free energy as a paradigm for scale bridging, we have previously exploited neural networks for their representation of such high-dimensional functions. Specifically, we have developed an integrable deep neural network (IDNN) that can be trained to free energy derivative data obtained from atomic scale models and statistical mechanics, then analytically integrated to recover a free energy density function. The motivation comes from the statistical mechanics formalism, in which certain free energy derivatives are accessible for control of the system, rather than the free energy itself. Our current work combines the IDNN with an active learning workflow to improve sampling of the free energy derivative data in a high-dimensional input space. Treated as input-output maps, machine learning accommodates role reversals between independent and dependent quantities as the mathematical descriptions change with scale bridging. As a prototypical system we focus on Ni-Al. Phase field simulations using the resulting IDNN representation for the free energy density of Ni-Al demonstrate that the appropriate physics of the material have been learned. To the best of our knowledge, this represents the most complete treatment of scale bridging, using the free energy for a practical materials system, that starts with electronic structure calculations and proceeds through statistical mechanics to continuum physics.

Convolutional neural networks (CNN) are utilized to encode the relation between initial configurations of obstacles and three fundamental quantities in porous media: porosity ($\varphi$), permeability $k$, and tortuosity ($T$). The two-dimensional systems with obstacles are considered. The fluid flow through a porous medium is simulated with the lattice Boltzmann method. It is demonstrated that the CNNs are able to predict the porosity, permeability, and tortuosity with good accuracy. With the usage of the CNN models, the relation between $T$ and $\varphi$ has been reproduced and compared with the empirical estimate. The analysis has been performed for the systems with $\varphi \in (0.37,0.99)$ which covers five orders of magnitude span for permeability $k \in (0.78, 2.1\times 10^5)$ and tortuosity $T \in (1.03,2.74)$.

Deep learning (DL) can achieve impressive results across a wide variety of tasks, but this often comes at the cost of training models for extensive periods on specialized hardware accelerators. This energy-intensive workload has seen immense growth in recent years. Machine learning (ML) may become a significant contributor to climate change if this exponential trend continues. If practitioners are aware of their energy and carbon footprint, then they may actively take steps to reduce it whenever possible. In this work, we present Carbontracker, a tool for tracking and predicting the energy and carbon footprint of training DL models. We propose that energy and carbon footprint of model development and training is reported alongside performance metrics using tools like Carbontracker. We hope this will promote responsible computing in ML and encourage research into energy-efficient deep neural networks.

This paper introduces a unified deep neural network (DNN)-based precoder for two-user multiple-input multiple-output (MIMO) networks with five objectives: data transmission, energy harvesting, simultaneous wireless information and power transfer, physical layer (PHY) security, and multicasting. First, a rotation-based precoding is developed to solve the above problems independently. Rotation-based precoding is new precoding and power allocation that beats existing solutions in PHY security and multicasting and is reliable in different antenna settings. Next, a DNN-based precoder is designed to unify the solution for all objectives. The proposed DNN concurrently learns the solutions given by conventional methods, i.e., analytical or rotation-based solutions. A binary vector is designed as an input feature to distinguish the objectives. Numerical results demonstrate that, compared to the conventional solutions, the proposed DNN-based precoder reduces on-the-fly computational complexity more than an order of magnitude while reaching near-optimal performance (99.45\% of the averaged optimal solutions). The new precoder is also more robust to the variations of the numbers of antennas at the receivers.