American tech giants are increasingly focusing on the humanoid robotics space, but analysts say they’re at risk of falling behind China.
Category: robotics/AI – Page 294
Advancing semiconductor devices for AI: Single transistor acts like neuron and synapse
Researchers from the National University of Singapore (NUS) have demonstrated that a single, standard silicon transistor, the fundamental building block of microchips used in computers, smartphones and almost every electronic system, can function like a biological neuron and synapse when operated in a specific, unconventional way.
Led by Associate Professor Mario Lanza from the Department of Materials Science and Engineering at the College of Design and Engineering, NUS, the research team’s work presents a highly scalable and energy-efficient solution for hardware-based artificial neural networks (ANNs).
This brings neuromorphic computing —where chips could process information more efficiently, much like the human brain —closer to reality. Their study was published in the journal Nature.
New miniature laboratories are ensuring that AI doesn’t make mistakes
Anyone who develops an AI solution sometimes goes on a journey into the unknown. At least at the beginning, researchers and designers do not always know whether their algorithms and AI models will work as expected or whether the AI will ultimately make mistakes.
Sometimes, AI applications that work well in theory perform poorly under real-life conditions. In order to gain the trust of users, however, an AI should work reliably and correctly. This applies just as much to popular chatbots as it does to AI tools in research.
Any new AI tool has to be tested thoroughly before it is deployed in the real world. However, testing in the real world can be an expensive, or even risky endeavor. For this reason, researchers often test their algorithms in computer simulations of reality. However, since simulations are approximations of reality, testing AI solutions in this way can lead researchers to overestimate an AI’s performance.
AI model transforms material design by predicting and explaining synthesizability
A research team has successfully developed a technology that utilizes Large Language Models (LLMs) to predict the synthesizability of novel materials and interpret the basis for such predictions. The team was led by Seoul National University’s Professor Yousung Jung and conducted in collaboration with Fordham University in the United States.
The findings of this research are expected to contribute to the novel material design process by filtering out material candidates with low synthesizability in advance or optimizing previously challenging-to-synthesize materials into more feasible forms.
The study, with Postdoctoral Researcher Seongmin Kim as the first author, was published in two chemistry journals: the Journal of the American Chemical Society on July 11, 2024, and Angewandte Chemie International Edition on February 13, 2025.
Wheel-less helical ring-based soft robot can move reliably in all directions
Over the past decades, roboticists have introduced a wide range of systems that can move in various complex environments, including different terrains, on the ground, in the air, and even in water. To safely navigate real-world dynamic environments without colliding with humans or nearby objects, most robots rely on sensors and cameras.
Researchers at Tsinghua University have recently developed WHERE-Bot, a new wheel-less, everting soft robot (i.e., a flexible robot that moves by turning its body structure inside out) that safely moves in unstructured environments without using sensors to detect obstacles. This robot, introduced in a paper published on the arXiv preprint server and set to be presented at the 8th IEEE International Conference on Soft Robotics (RoboSoft) in April, leverages its unique helical ring-based structure to move in all directions.
“One day, while playing with a Slinky toy during a lab meeting,” Shuguang Li, senior author of the paper, told Tech Xplore. “Suddenly, a new idea struck us: what if we connected the head and tail of the spring toy? By joining its two ends, the spring could be endlessly turned inside-out—a motion we now call ‘everting’—presenting a fascinating color flow. This sparked our curiosity about how such a helical ring—perhaps with some structure modifications—would behave in various environments: on the ground, along a pipe, underwater, on sand, and even in the air.”
Dirac’s Plate Trick, the Hairy Ball Theorem and more: Research probes physics of irregular objects on inclined planes
How gravity causes a perfectly spherical ball to roll down an inclined plane is part of the elementary school physics canon. But the world is messier than a textbook.
Scientists in the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) have sought to quantitatively describe the much more complex rolling physics of real-world objects. Led by L. Mahadevan, the Lola England de Valpine Professor of Applied Mathematics, Physics, and Organismic and Evolutionary Biology in SEAS and FAS, they combined theory, simulations, and experiments to understand what happens when an imperfect, spherical object is placed on an inclined plane.
Published in Proceedings of the National Academy of Sciences, the research, which was inspired by nothing more than curiosity about the everyday world, could provide fundamental insights into anything that involves irregular objects that roll, from nanoscale cellular transport to robotics.
Liquid-crystal platform overcomes optical losses in photonic circuits
Photonic circuits, which manipulate light to perform various computational tasks, have become essential tools for a range of advanced technologies—from quantum simulations to artificial intelligence. These circuits offer a promising way to process information with minimal energy loss, especially in fields like quantum computing where complex systems are simulated to test theories of quantum mechanics.
However, the growth in circuit size and complexity has historically led to a rise in optical losses, making it challenging to scale these systems for large-scale applications, such as multiphoton quantum experiments or all-optical AI systems.
As reported in Advanced Photonics, researchers at the University of Naples Federico II have now developed a new approach to address this problem. Using a liquid-crystal (LC)-based platform, the team designed an optical processor capable of handling hundreds of optical modes in a compact, two-dimensional setup. This breakthrough offers a solution to a key limitation in traditional photonic circuits, where losses increase as the number of modes grows.
Who’s to blame when AI makes a medical error?
Assistive artificial intelligence technologies hold significant promise for transforming health care by aiding physicians in diagnosing, managing, and treating patients. However, the current trend of assistive AI implementation could actually worsen challenges related to error prevention and physician burnout, according to a new brief published in JAMA Health Forum.
The brief, written by researchers from the Johns Hopkins Carey Business School, Johns Hopkins Medicine, and the University of Texas at Austin McCombs School of Business, explains that there is an increasing expectation of physicians to rely on AI to minimize medical errors. However, proper laws and regulations are not yet in place to support physicians as they make AI-guided decisions, despite the fierce adoption of these technologies among health care organizations.
The researchers predict that medical liability will depend on whom society considers at fault when the technology fails or makes a mistake, subjecting physicians to an unrealistic expectation of knowing when to override or trust AI. The authors warn that such an expectation could increase the risk of burnout and even errors among physicians.