The Path Planning Transformer (PPT) model learns to plan efficient paths from occupancy maps, avoiding obstacles with a modified right-of-way rule. This approach improves path smoothness and adaptability while reducing computational requirements, with potential applications in industrial automation and collaborative robot systems.
A new study provides a refined framework for understanding biotic evolution immediately preceding the Late Ordovician Mass Extinction in South China. The study reveals diverse shallow marine biotas in the region and reconstructs the regional ecosystem with high resolution.
Researchers propose a multi-channel TDC with Physically Unclonable Function (PUF) to generate unique fingerprints of ICs. This approach addresses security concerns in programmable devices and offers improved performance and lower cost.
Advanced Cancer Research defines a platform for integrating molecular biology, genomics, immunology, engineering, and computational science to understand cancer as a complex disease. The journal prioritizes studies that provide mechanistic depth and bridge disciplinary boundaries.
Researchers developed a highly accurate computer model to simulate the entire tunnel excavation process, comparing ground anchors and pre-support pipes. The study found that a combined lining and anchor system is sufficient to control rock stress and deformation in deep mountain tunnels.
InstaDrive generates precise editing of vehicles and map elements, enabling efficient labeled data generation. It outperforms baselines in FID and mAP, preserving accurate map structures and maintaining multi-view consistency.
Researchers investigate real-world blockchain security incidents to extract representative attack patterns and categorize them into six classes based on standard blockchain architecture layers. The study also evaluates current detection and defense strategies, highlighting their strengths and limitations.
Researchers developed Zr-IR825 nanoparticles by combining IR825 with biocompatible metal zirconium, improving water solubility and stability. The material exhibited outstanding photothermal conversion capabilities under near-infrared light, effectively killing tumor cells while showing low toxicity to normal cells.
Researchers developed PDJA, a novel AI-based attack framework that amplifies vulnerability and bypasses defenses in multi-agent reinforcement learning systems. The approach improves attack effectiveness and cross-layer vulnerability exploitation, opening new opportunities for evaluating the robustness of AI-driven autonomous systems.
A new study assesses the EU Energy Performance of Buildings Directive through a social justice lens, highlighting how renovation policies can widen existing social inequalities. The Composite Building and Social Vulnerability Index reveals markedly different levels of vulnerability across EU countries.
Researchers developed an AI–DFT integrated framework to accelerate materials discovery and design, enabling the efficient exploration of vast chemical and structural spaces. The framework combines the accuracy of first-principles calculations with the speed of AI models, overcoming data scarcity and computational cost challenges.
Researchers developed a new 8B13B coding scheme to enable robust, low-cost visible light communication using commercially available components. The system achieved reliable outdoor VLC at data rates of up to 3.48 Mbit/s over distances of several meters, overcoming key challenges in VLC.
Researchers explore how nanomaterials enhance drug stability, delivery, and medical imaging in treating eye tumors. The concept of theranostics combines diagnosis and therapy for personalized treatments.
Researchers developed a hierarchical Physics-Informed AI framework to balance model fidelity with speed, paving the way for reliable Digital Twins in solid oxide cell engineering. The new approach improves both physical consistency and computational efficiency.
Researchers developed a physics-constrained, data-efficient AI framework that accurately reconstructs temperature fields from limited sensor data in chiplet-based packaging. The approach achieves high accuracy and generalization without excessive noise, enabling reliable thermal characterization.
Researchers developed an AI-driven framework using XGBoost to evaluate asphalt pavement skid resistance under various conditions, achieving over 90% prediction accuracy. The model identified fractal dimension as the core predictive factor, highlighting the role of texture complexity in skid resistance.
A new study presents a comprehensive multi-scale computational framework capable of predicting the mechanical responses of Fe–Cr–Al alloys. The framework integrates molecular dynamics, phase-field simulations, and finite element modelling to simulate solid-solution effects and processing-induced microstructural evolution.
Researchers have developed a method to fabricate multilayer optical coatings using natural mineral powders, reducing material costs and environmental impact. The use of raw materials like SiO2, TiO2, and iron/copper oxides enables the creation of vibrant colors with high durability and resistance.
Researchers developed a socially aware prediction-to-control pipeline to enable autonomous vehicles to safely weave through dense crowds. The integrated framework achieved zero safety violations and maintained comfortable motion while meeting real-time computing limits.
Researchers introduce HierSpectrumChain, a hierarchical blockchain architecture for efficient and secure dynamic spectrum sharing among diverse wireless users. The system automates spectrum leasing through smart-contract–driven Stackelberg auctions, minimizing latency while maintaining transparency and trust.
The conference will bring together thousands of experts to share research and attend sessions on the latest Computer, Big Data and Artificial Intelligence developments. Authors can submit full research papers and benefit from a registration fee waiver for selected outstanding papers.
A new study benchmarks six prompt styles across three leading LLMs to solve the challenge of balancing accuracy, speed, and cost in structured data workflows. The research reveals distinct strengths in each model, highlighting the need for tailored approaches for different data types.
The POSTECH team proposes two innovative strategies using nanoimprint lithography to address the limitations of traditional metasurface manufacturing. The hybrid material method and particle-embedded resin method enable high-performance metasurfaces with optical efficiencies comparable to electron beam lithography, making large-scale m...
A new nonlinear dynamic modeling framework captures complex tower-blade interactions, including torsional effects, achieving a remarkable agreement with high-fidelity benchmarks. The framework enables accurate and efficient simulations, contributing to the design of lighter, safer, and more economically competitive wind turbine towers.
Researchers developed an AI-based method to process high-frequency welding data, achieving a 28.3% improvement in anomaly detection performance. The approach uses time-frequency domain analysis and Isolation Forest to capture structural patterns underlying the repetitive nature of the welding process.
Omni-modal language models integrate perception, understanding, and reasoning within a unified framework, bringing AI closer to human cognition. Lightweight adaptation strategies improve deployment efficiency in real-time medical and industrial scenarios.
The GBCESC 2025 conference aims to exchange technical knowledge and innovative solutions in green building, civil engineering, and smart city fields. Accepted papers will be published in the conference proceedings.
The 2nd International Conference on Civil Engineering and Smart Construction (ICCESC 2025) aims to bridge academic accomplishments with industrial progress. The conference covers various areas of interest including civil engineering, smart construction, and intelligent technologies.
Anticancer nanoparticles pose risks from drug payloads, nanomaterial accumulation, and bio-corona formation. Measures to improve safety include biodegradable materials, surface engineering, and organ-specific delivery systems.
Researchers have developed a lightweight energy-absorbing box using PEEK honeycomb, increasing average crash load by 18.1% and offering superior balance between safety and weight reduction.
Researchers have developed machine learning models to accelerate the discovery and design of efficient 2D electrocatalysts, enabling faster and more reliable catalyst optimization. The study highlights the importance of high-quality data, descriptor selection, and interpretable modeling in harnessing the full potential of machine learn...
Researchers propose a novel, data-driven solution using machine learning to predict optimal nanocarrier design for individual patients' tumors. The approach enhances synchronization between drug release and peak tumor proliferation rates, increasing its by a factor of 2.8 compared to traditional static nanocarriers.
The LiteRBS algorithm outperforms classical algorithms like A* and Bidirectional A*, scaling well even in large or dense maps. It achieves fast and memory-efficient pathfinding through an aggressive bidirectional forward search with a reserve-queue fallback strategy.
Researchers have developed an AI method to control and stabilize the Uzawa-Lucas endogenous growth model, which describes interaction between physical capital and human capital. The flatness-based adaptive fuzzy controller uses partial data and leverages mathematical properties to ensure global stability.
The study highlights how wind farms have transformed rural landscapes, disrupting local agricultural markets and eroding rural identities. The research calls for a more participatory and place-based energy transition that respects local heritage and democratic governance.
A new study published in AIAS found that warm red lighting can significantly reduce motion sickness in passengers during nighttime journeys. The research team conducted nighttime road tests and combined subjective ratings with multimodal physiological signals to assess symptoms.
Researchers developed nacre-derived biphasic calcium phosphate composite scaffolds that combine osteogenic and angiogenic functions, accelerating bone regeneration and promoting vascular growth. These 'smart scaffolds' show great promise for efficient bone defect repair and could serve as a bone graft substitute.
A new smart tracking system uses barometric sensors and LoRaWAN to measure height and identify materials on multi-story construction sites. The system has been tested in controlled environments and shows promise for improving material tracking and reducing waste and costs.
Researchers developed a robust QA system using Retrieval Augmented Generation (RAG) to answer queries on building codes. They found that Elasticsearch was the most effective retriever and fine-tuning Large Language Models (LLMs) significantly enhanced generational accuracy, achieving a 6.83% relative improvement in BERT F1-score.
A computational study reveals polyethylene furanoate (PEF) has stronger resistance to hydrogen permeation than PA6 and PE, thanks to its rigid furan ring structure. PEF's superior barrier performance makes it a promising alternative for high-pressure hydrogen storage liners.
A new review in AI & Materials highlights the transformative potential of machine learning (ML) in nuclear materials research. ML techniques are being used to analyze complex microstructures, predict thermal conductivity and mechanical behavior, optimize processing and fabrication, and integrate with physics-based models.
Researchers from Shandong Technology and Business University survey the latest AI-powered techniques to fill in missing traffic data gaps. The study categorizes and compares leading data imputation methods, offering a clear roadmap for researchers and city planners to improve traffic management and smart city operations.
The study combines decision making and machine learning to improve robot learning for water analysis. The Random Forest Classifier achieved an accuracy of 69% which improved to 73% after applying Synthetic Minority Over-sampling Technique (SMOTE) and tuning hyperparameters. The robotic system can detect, analyze, and distinguish drinki...
Researchers have created a novel method to culture living brain cells in Alzheimer's and Parkinson's disease, using affordable 'ring-inserts' that support high-quality tissue cultures and enable real-time imaging. This cost-effective approach reduces experimental costs by up to €3,500 while maintaining experimental quality.
The study assesses autonomous vehicle adaptability on small-radius circular curves, showing ride comfort degrades to
Researchers propose a novel control scheme to address challenges in traditional backstepping control designs and event-triggered control. The scheme significantly reduces communication bandwidth usage, offering a new approach for efficient control of complex industrial systems.
Researchers developed an advanced modeling framework to enhance point cloud quality in metro tunnel inspections using backpack SLAM LiDAR systems. Inspection speed and scan density emerged as crucial determinants of point cloud quality, with optimal operational conditions achieving high-quality data.
A new framework leverages automated 3D object segmentation and mesh reconstruction to simulate dynamic and realistic construction sites. The approach captures real-world scenes, extracts key objects, and reconstructs them into editable 3D meshes, enabling applications such as virtual safety walkthroughs and AI-driven layout planning.
Researchers explore emerging role of extracellular RNAs in cancer diagnosis, prognosis, and treatment. These natural nanocarriers can reflect tumor behavior in real time and deliver targeted therapies.
The research develops a comprehensive family assistant capable of performing various tasks like picking up toys and fetching items. The robot uses YOLOv11 object recognition and flexible gripper to recognize objects of varying sizes and weights with high accuracy.