Conventional machine learning approaches accelerate inorganic materials design via accurate property prediction and targeted material generation, yet they operate as single-shot models limited by the latent knowledge baked into their training data. A central challenge lies in creating an intelligent system capable of autonomously executing the full inorganic materials discovery cycle, from ideation and planning to experimentation and iterative refinement. We introduce SparksMatter, a multi-agent AI model for automated inorganic materials design that addresses user queries by generating ideas, designing and executing experimental workflows, continuously evaluating and refining results, and ultimately proposing candidate materials that meet the target objectives. SparksMatter also critiques and improves its own responses, identifies research gaps and limitations, and suggests rigorous follow-up validation steps, including DFT calculations and experimental synthesis and characterization, embedded in a well-structured final report. The model's performance is evaluated across case studies in thermoelectrics, semiconductors, and perovskite oxides materials design. The results demonstrate the capacity of SparksMatter to generate novel stable inorganic structures that target the user's needs. Benchmarking against frontier models reveals that SparksMatter consistently achieves higher scores in relevance, novelty, and scientific rigor, with a significant improvement in novelty across multiple real-world design tasks as assessed by a blinded evaluator. These results demonstrate SparksMatter's unique capacity to generate chemically valid, physically meaningful, and creative inorganic materials hypotheses beyond existing materials knowledge.

Microplastics, defined as plastic particles smaller than 5 millimeters, have become a pervasive environmental contaminant that accumulates on beaches due to wind patterns and tidal forcing. Detecting microplastics and mapping their concentration in the wild remains one of the primary challenges in addressing this environmental issue. This paper introduces a novel robotic platform that automatically detects and chemically analyzes microplastics on beach surfaces. This mobile manipulator system scans areas for microplastics using a camera mounted on the robotic arm's end effector. The system effectively segments candidate microplastic particles on sand surfaces even in the presence of organic matter such as leaves and clams. Once a candidate microplastic particle is detected, the system steers a near-infrared (NIR) spectroscopic sensor onto the particle using both NIR and visual feedback to chemically analyze it in real-time. Through experiments in lab and beach environments, the system is shown to achieve an excellent positional precision in manipulation control and high microplastic classification accuracy.

--Edge Digital Twins (EDTs) are crucial for monitoring and control of Power Electronics Systems (PES). However, existing modeling approaches struggle to consistently capture continuously evolving hybrid dynamics that are inherent in PES, degrading Sim-to-Real generalization on resource-constrained edge devices. To address these challenges, this paper proposes a Physics-Embedded Neural ODEs (PENODE) that (i) embeds the hybrid operating mechanism as an event automaton to explicitly govern discrete switching and (ii) injects known governing ODE components directly into the neural parameterization of unmodeled dynamics. This unified design yields a differentiable end-to-end trainable architecture that preserves physical interpretability while reducing redundancy, and it supports a cloud-to-edge toolchain for efficient FPGA deployment. Experimental results demonstrate that PENODE achieves significantly higher accuracy in benchmarks in white-box, gray-box, and black-box scenarios, with a 75% reduction in neuron count, validating that the proposed PENODE maintains physical interpretability, efficient edge deployment, and real-time control enhancement. OWER electronics systems (PES) are the fundamental to drive efficient energy conversion [1] but require precise and real-time monitoring and predictive analysis due to the ultra-high standards for reliability and performance [2]. Digital Twin (DT), a high-fidelity virtual counterpart of a physical asset, presents a promising solution [3]. However, it is difficult to implement cloud-or-server-based DTs with high communication latency and limited bandwidth in PES because the PES dynamics differ significantly from power grids [4].

In 2012, AlexNet established deep convolutional neural networks (DCNNs) as the state-of-the-art in CV, as these networks soon led in visual tasks for many domains, including remote sensing. With the publication of Visual Transformers, we are witnessing the second modern leap in computational vision, and as such, it is imperative to understand how various transformer-based neural networks perform on satellite imagery. While transformers have shown high levels of performance in natural language processing and CV applications, they have yet to be compared on a large scale to modern remote sensing data. In this paper, we explore the use of transformer-based neural networks for object detection in high-resolution electro-optical satellite imagery, demonstrating state-of-the-art performance on a variety of publicly available benchmark data sets. We compare eleven distinct bounding-box detection and localization algorithms in this study, of which seven were published since 2020, and all eleven since 2015. The performance of five transformer-based architectures is compared with six convolutional networks on three state-of-the-art open-source high-resolution remote sensing imagery datasets ranging in size and complexity. Following the training and evaluation of thirty-three deep neural models, we then discuss and analyze model performance across various feature extraction methodologies and detection algorithms. Machine learning and computer vision (CV) have seen the incredible rise of deep neural networks (DNNs), particularly convolutional neural networks, since the original AlexNet [1] paper in 2012. Combined with the processing power of GPUs and the increasing availability of robust pre-trained weights derived from massive image-sets for techniques like transfer learning, the ability to effectively train deep models has in turn led to DNNs becoming the most widely used CV technique. In recent years, however, the convolutional feature extractors that have long been the foundation of these DNNs have been outperformed on CV challenge datasets such as the ImageNet [2] and COCO [3] competitions by a newer feature extraction architecture, known as transformers. Convolutional-based deep neural networks have historically shown outstanding performance in CV applications, however, the recent publication of Visual Transformer Neural Networks, beginning with the original Vision Transformer (ViT) [4], has enabled a leap in computational vision capabilities. Following the publication of ViT, visual transformer architectures have been found to be capable of outperforming traditional convolutional networks for a variety of CV applications.

Wearable assistive devices are increasingly becoming softer. Modelling their interface with human tissue is necessary to capture transmission of dynamic assistance. However, their nonlinear and compliant nature makes both physical modeling and embedded sensing challenging. In this paper, we develop a image-based, learning-based framework to estimate distributed contact forces for a finger-exosuit system. We used the SoRoSim toolbox to generate a diverse dataset of exosuit geometries and actuation scenarios for training. The method accurately estimated interaction forces across multiple contact locations from low-resolution grayscale images, was able to generalize to unseen shapes and actuation levels, and remained robust under visual noise and contrast variations. We integrated the model into a feedback controller, and found that the vision-based estimator functions as a surrogate force sensor for closed-loop control. This approach could be used as a non-intrusive alternative for real-time force estimation for exosuits.

Since the dawn of civilization, humans have tried to understand the nature of intelligence. With the advent of computers, there have been attempts to emulate human intelligence using computer algorithms - a field that was dubbed'Artificial Intelligence' or'AI' by the computer scientist John McCarthy in 1956 and has recently enjoyed an explosion of popularity. Many efforts in AI research have focused on the study and replication of what is considered the hallmark of human cognition, such as playing intelligent games, the faculty of language, visual perception, and creativity. While at the time of writing we have multiple successful takes at the above - computers nowadays play chess and Go better than any human, can translate English into Chinese without a dictionary, automatically drive a car in a crowded city, and generate poetry and art that wins artistic competitions - it is fair to say that we still do not have a full understanding of what human-like or'general' intelligence entails and how to replicate it.

Prior models and benchmarks focus on closed-loop motion simulation for initial agents in a scene. This is problematic for long-term simulation. Agents enter and exit the scene as the ego vehicle enters new regions. W e propose InfGen, a unified next-token prediction model that performs interleaved closed-loop motion simulation and scene generation. InfGen automatically switches between closed-loop motion simulation and scene generation mode. It enables stable long-term rollout simulation. InfGen performs at the state-of-the-art in short-term (9s) traffic simulation, and significantly outperforms all other methods in long-term (30s) simulation.

This paper proposes a novel theoretical framework for guaranteeing and evaluating the resilience of long short-term memory (LSTM) networks in control systems. We introduce "recovery time" as a new metric of resilience in order to quantify the time required for an LSTM to return to its normal state after anomalous inputs. By mathematically refining incremental input-to-state stability ($δ$ISS) theory for LSTM, we derive a practical data-independent upper bound on recovery time. This upper bound gives us resilience-aware training. Experimental validation on simple models demonstrates the effectiveness of our resilience estimation and control methods, enhancing a foundation for rigorous quality assurance in safety-critical AI applications.

A campground is expanding with 43 new hotel rooms and 18 homes, all built by a massive 3D printer. A remarkable new home in Japan is turning heads and turning the construction industry on its ear. Known as the Lib Earth House Model B, this single-story home was created using 3D-printing technology and a soil-based mixture instead of traditional concrete. It's a bold move toward sustainability, blending innovation with nature in a way that could redefine how we build homes around the world. Sign up for my FREE CyberGuy Report Get my best tech tips, urgent security alerts, and exclusive deals delivered straight to your inbox.

Machine learning models trained on Earth observation data, such as satellite imagery, have demonstrated significant promise in predicting household-level wealth indices, enabling the creation of high-resolution wealth maps that can be leveraged across multiple causal trials. However, because standard training objectives prioritize overall predictive accuracy, these predictions inherently suffer from shrinkage toward the mean, leading to attenuated estimates of causal treatment effects and limiting their utility in policy. Existing debiasing methods, such as Prediction-Powered Inference, can handle this attenuation bias but require additional fresh ground-truth data at the downstream stage of causal inference, which restricts their applicability in data-scarce environments. Here, we introduce and evaluate two correction methods -- linear calibration correction and Tweedie's correction -- that substantially reduce prediction bias without relying on newly collected labeled data. Linear calibration corrects bias through a straightforward linear transformation derived from held-out calibration data, whereas Tweedie's correction leverages empirical Bayes principles to directly address shrinkage-induced biases by exploiting score functions derived from the model's learning patterns. Through analytical exercises and experiments using Demographic and Health Survey data, we demonstrate that the proposed methods meet or outperform existing approaches that either require (a) adjustments to training pipelines or (b) additional labeled data. These approaches may represent a promising avenue for improving the reliability of causal inference when direct outcome measures are limited or unavailable, enabling a "one map, many trials" paradigm where a single upstream data creation team produces predictions usable by many downstream teams across diverse ML pipelines.