At GE Research, we are combining “physics” with artificial intelligence and machine learning to advance manufacturing design, processing, and inspection, turning innovative technologies into real products and solutions across our industrial portfolio. This article provides a snapshot of how this physical plus digital transformation is evolving at GE.

Atomically resolved imaging of materials enabled by the advent of aberration-corrected scanning transmission electron microscopy (STEM) has become a mainstay of modern materials science. However, much of the wealth of quantitative information contained in the fine details of atomic structure or spectra remains largely unexplored. In this article, we discuss new opportunities enabled by physics-informed big data and machine learning technologies to extract physical information from static and dynamic STEM images, ranging from statistical thermodynamics of alloys to kinetics of solid-state reactions at a single defect level. The synergy of deep-learning image analytics and real-time feedback further allows harnessing beam-induced atomic and bond dynamics to enable direct atom-by-atom fabrication. Examples of direct atomic motion over mesoscopic distances, engineered doping at selected lattice sites, and assembly of multiatomic structures are reviewed. These advances position the scanning transmission electron microscope to transition from a mere imaging tool toward a complete nanoscale laboratory for exploring electronic, phonon, and quantum phenomena in atomically engineered structures.

A fundamental design rule that nature has developed for biological machines is the intimate correlation between motion and function. One class of biological machines is molecular motors in living cells, which directly convert chemical energy into mechanical work. They coexist in every eukaryotic cell, but differ in their types of motion, the filaments they bind to, the cargos they carry, as well as the work they perform. Such natural structures offer inspiration and blueprints for constructing DNA-assembled artificial systems, which mimic their functionality. In this article, we describe two groups of cytoskeletal motors, linear and rotary motors. We discuss how their artificial analogues can be built using DNA nanotechnology. Finally, we summarize ongoing research directions and conclude that DNA origami has a bright future ahead.

Discovering knowledge from data is a quantum jump from quantity to quality, which is the characteristic and the spirit of the development of science. Symbolic regression (SR) is playing a greater role in the discovery of knowledge from data, specifically in this era of exponential data growth, because SRs are able to discover mathematical formulas from data. These formulas may provide scientifically meaningful models, especially when combined with domain knowledge. This article provides an overview of SR applications in the field of materials science and engineering. Integrating domain knowledge with SR is the key and a crucial approach, which allows gaining knowledge from data quickly, accurately, and scientifically. In the data-driven paradigm, SR allows for uncovering the underlying mechanisms of materials behavior, properties, and functions, in a wide range of areas from basic academic research to industrial applications, including experiments and computations, by providing explicit interpretable models from data, in comparison with other machine-learning “black-box” models. SR will be a powerful tool for rational and automatic materials development.

Continued progress in artificial intelligence (AI) and associated demonstrations of superhuman performance have raised the expectation that AI can revolutionize scientific discovery in general and materials science specifically. We illustrate the success of machine learning (ML) algorithms in tasks ranging from machine vision to game playing and describe how existing algorithms can also be impactful in materials science, while noting key limitations for accelerating materials discovery. Issues of data scarcity and the combinatorial nature of materials spaces, which limit application of ML techniques in materials science, can be overcome by exploiting the rich scientific knowledge from physics and chemistry using additional AI techniques such as reasoning, planning, and knowledge representation. The integration of these techniques in materials-intelligent systems will enable AI governance of the scientific method and autonomous scientific discovery.