Adaptive Systems
Articles tagged with Adaptive Systems
Machine learning proves that graphene is hydrophobic
Researchers used machine-learning-enhanced molecular simulations to show pristine graphene is intrinsically hydrophobic. Water molecules adopt configurations characteristic of hydrophobic surfaces near graphene, and thicker layers are even more strongly hydrophobic.
WVU legal expert finds judges cautiously adopting AI while guarding human authority
New research from West Virginia University finds that judges are adopting generative artificial intelligence in courtrooms, but remain committed to human control over judicial decision-making. Judges use AI for administrative tasks like document summarization and case organization, but prioritize legal reasoning and final judgment.
AI species evolving like organisms may soon emerge – and create risks
Researchers warn of potential risks associated with evolvable AI systems, which can tap into the power of biological evolution to create 'selfish' actors that break alignment with human goals. The study recommends guardrails to maintain centralized control over AI reproduction and mitigate risks.
AI models lean on autism stereotypes when giving social advice, new study finds
Artificial intelligence models provide personalized advice, but may perpetuate negative stereotypes about people with autism. Researchers found that up to 70% of the time, AI discourages those with autism from socializing.
Machine learning designs cheaper and rust-proof steel for 3D printing
A new class of ultra-high strength and ductility steel has been created using machine learning, achieving a rare balance of extreme strength and ductility. The resulting metal resists corrosion and degrades slowly in salt-water tests.
Eureka! Scientists develop new way to detect breakthroughs in science
A team of researchers at Binghamton University has developed a method to pinpoint discoveries that reshaped the course of science. The new metric uses neural embedding to analyze approximately 55 million scientific papers and patents, identifying major breakthroughs and simultaneous discoveries with greater accuracy.
New tool maps the landscape of student knowledge using short quizzes
Researchers at Dartmouth College developed a mathematical framework to map students' conceptual knowledge from short multiple-choice quizzes, revealing peaks of mastery and valleys of struggle. The technique could enable personalized learning, AI tutoring systems, and more efficient feedback.
Want to shift a group’s opinion? Encourage opponents to sit on the fence
Researchers propose a strategy that encourages individuals to adopt a neutral stance, allowing groups to become more responsive, decisions to become easier to reach, and shifts in consensus to happen smoothly. By doing so, neutrality creates valuable breathing space for reassessment, making it easier for a group to change its mind when...
Ateneo machine learning lab opens doors to industry partners, collaborators
The Ateneo Laboratory for Intelligent Visual Environments (ALIVE) is developing machine learning solutions with industry partners to improve public health, traffic systems, and more. By bridging the gap between messy reality and mathematical models, ALIVE is creating intelligent visual systems that can handle real-world conditions.
The brain’s primitive ‘fear center’ is actually a sophisticated mediator
A Dartmouth study challenges the conventional view of the amygdala as a primitive 'fear center' by revealing its role in mediating between competing learning strategies. The research suggests that the amygdala favors action-based learning, promoting exploration and flexibility to overcome fear.
Exposing biases, moods, personalities, and abstract concepts hidden in LLMs
A team from MIT and UC San Diego has developed a new method to uncover hidden biases, moods, and abstract concepts in large language models (LLMs). The approach identifies these connections within the model and allows for manipulation of the concept in generated answers.
Using the physics of radio waves to empower smarter edge devices
Researchers at Duke University have created a new method to use analog radio waves to boost energy-efficient edge AI, enabling devices to run powerful AI models without heavy chips or distant servers. The approach, called Wireless Smart Edge networks (WISE), achieves nearly 96% image classification accuracy while consuming significantl...
Revolutionizing biosecurity: new multi-omics framework to transform invasive species management
A new multi-omics framework proposes a proactive, predictive, and integrative approach to invasive species management. The framework uses advanced technologies to detect, track, and manage invasive species with unprecedented precision.
Colorectal cancer survival predicted by AI using clinical and molecular features
Researchers used machine learning to combine clinical data with biological markers, achieving 89.58% accuracy in predicting patient survival. The study identified key markers such as E2F8, WDR77, and hsa-miR-495-3p, which were associated with tumor growth and cancer development.
Frontiers in Science Deep Dive series: How breaking the ‘memory wall’ using brain-inspired algorithms could help overcome AI energy costs
Researchers propose a novel approach to AI hardware design by integrating neuromorphic systems and compute-in-memory techniques to overcome the limitations of modern computing hardware. This could lead to more efficient data center energy use and enable real-time intelligence in compact, power-constrained systems.
How brain-inspired algorithms could drive down AI energy costs
Researchers propose integrating processing capability within memory units to reduce energy consumption and latency in AI applications. Inspired by the brain's efficient processing mechanisms, spiking neural networks (SNNs) can respond to irregular events and store information in the same place.
Can AI read humans’ minds? A new model shows it’s shockingly good at it
A breakthrough AI system called OmniPredict can predict human pedestrian behaviors with unprecedented accuracy, revolutionizing self-driving cars and urban mobility. The model combines visual cues with contextual information to anticipate pedestrians' next moves, reducing the risk of accidents and improving traffic safety.
AI can deliver personalized learning at scale, study shows
A Dartmouth study finds that AI-powered chatbots can deliver personalized learning to large numbers of students. The researchers created an AI teaching assistant called NeuroBot TA that provides around-the-clock individualized support for students, which they found to be more trusted than general chatbots.
Lehigh University–Siemens partnership advances microgrid research for AI data centers
A PhD student at Lehigh University is working with Siemens to develop real-time monitoring and control tools for hyperscale data centers. The goal is to create a localized power network that can operate independently of the main grid, reducing power demands from artificial intelligence and increasing energy efficiency.
Drug toxicity predicted by differences between preclinical models and humans
A new AI model uses machine learning to predict drug toxicity in humans by identifying biological differences between cells, mice, and humans. The model improved predictive power over existing state-of-the-art models and demonstrated practicality in predicting market withdrawal due to toxicity.
Researchers pose five guiding questions to improve the use of artificial intelligence in physicians’ clinical decision-making
A research team provides a framework to support doctors in their patient care while ensuring AI doesn't undermine their expertise. The framework addresses key issues like timing, trust, and over-reliance on AI.
How can (A)I help you?
A new study by Yifan Yu offers guidance on how to deploy emotion AI in various scenarios, emphasizing the importance of balancing human involvement with AI's emotional detection capabilities. The analysis showed that emotion AI works best when integrated with human employees, and some scenarios are better handled by humans alone.
Hitting a nerve
Engineers at the University of Pittsburgh have created a soft material with a nerve net that mimics how simple living systems coordinate motion. The material responds to chemical reactions, producing mechanical movement without electronics or motors.
Generative AI enters nephrology: Towards an augmented medicine
The conference explores how generative AI is reshaping kidney medicine through AI-driven diagnostics, data integration, and omics analysis. Key findings include the use of LLMs to transform diagnostic precision, patient management, and research design.
Metal, melted, mastered
Researchers at Virginia Tech have developed an AI-powered system to detect flaws in wire-arc additive manufacturing, a faster approach to producing complex components. The technology enables real-time defect detection and correction, reducing waste and improving quality.
HealthFORCE, AAPA, and West Health release “Aging Well with AI” – first in a two part series on AI and the healthcare workforce
A new report by HealthFORCE, AAPA, and West Health highlights five ways AI can reduce strain on clinicians and improve outcomes for older adults. The paper aims to strengthen the US healthcare workforce and improve access to care as the nation confronts a historic shortage of healthcare workers alongside a rapidly aging population.
MoBluRF: A framework for creating sharp 4D reconstructions from blurry videos
Researchers developed MoBluRF, a two-stage motion deblurring method for NeRFs, achieving high-quality 3D reconstructions from ordinary blurry videos. The framework outperforms state-of-the-art methods and is robust against varying degrees of blur, enabling smartphones to produce sharper and more immersive content.
New book explores agent-based modeling, multi-agent systems
The book provides an overview of agent-based modeling and multi-agent systems, highlighting their application in understanding economic crises. It integrates machine learning to enhance adaptation and behavior of agents in dynamic environments.
Machine learning accelerates biochar research to cut carbon emissions and recycle waste
Biochar, a carbon-rich material, is gaining attention for its ability to improve soils, clean water, and capture carbon. Machine learning models can predict biochar yield and pollutant removal efficiency with over 90% accuracy, accelerating its development.
WVU researchers train AI to diagnose heart failure in rural patients using low-tech electrocardiograms
Researchers developed AI models that can identify signs of heart failure in patients from Appalachia using low-tech electrocardiogram results. The models achieved high accuracy and could potentially provide clinicians with an edge in protecting patients' cardiac health.
Smarter robot planning for the real world
Vasile's research aims to map and model an agent's capabilities, particularly in motion, manipulation, and perception, to reliably predict their behavior. The goal is to use this understanding to plan effectively for large teams of agents.
First transfer of behavior between species through single gene manipulation
Researchers successfully transferred a gift-giving courtship behavior from Drosophila subobscura to Drosophila melanogaster by manipulating a single gene in insulin-producing neurons. This study represents the first example of transferring behavior between species through genetic manipulation.
New book highlights real-world applications of artificial intelligence (AI) and machine learning (ML)
The new book highlights the transformative role of artificial intelligence (AI) and machine learning (ML) across various domains, including mechatronics, cybersecurity, digital health, and automation. Readers will gain practical insights into AI-based techniques in power systems, social media management, and healthcare diagnostics.
Breakthrough robotic slip-prevention method could bring human-like dexterity to industrial automation
A new robotic slip-prevention method has been developed to improve robots' grip and handling of fragile or slippery objects. This bio-inspired approach allows robots to predict when an object might slip and adapt their movements in real-time, outperforming traditional strategies.
Animal-inspired AI robot learns to navigate unfamiliar terrain
Researchers developed an AI system that enables a four-legged robot to adapt its gait to different terrain, just like animals. The robot learned to switch gaits on the fly and navigate uneven surfaces without any alterations to the system itself, overcoming previous limitations around adaptability.
AI medical receptionist modernizing doctor appointments, poised to improve patient care nationwide
Cassie, a digital-human assistant developed by Texas A&M University, is transforming the way patients interact with healthcare providers. With facial recognition and emotional intelligence, Cassie offers a two-way interaction that feels like a conversation.
Soft robot modules for new haptic interactions
The Digits framework uses compressed air to produce shape changes, vibrations, and haptic feedback, offering a versatile platform for virtual reality and physical therapy. The device's modular design and pneumatic actuation enable adaptable and scalable control methods.
New research finds specific learning strategies can enhance AI model effectiveness in hospitals
Researchers used proactive and transfer learning strategies to mitigate data shifts in AI models for hospital applications. They found that models trained on one hospital type performed better than those trained on all hospitals using transfer learning.
Engineering smarter drones: From nature to complex aerial manipulation
Aerial robots are limited to manipulating rigid objects, but Lehigh University researcher David Saldaña aims to expand their capabilities with an adaptive controller and reinforcement learning. His research has potential applications in construction, disaster response, and industrial automation.
Machine learning simplifies industrial laser processes
Researchers from Empa developed machine learning algorithms to optimize laser-based manufacturing techniques, reducing preliminary experiments by two-thirds. They also implemented real-time optimization using field-programmable gate arrays (FPGAs) for improved welding processes.
WVU researchers test AI’s limits in emergency room diagnoses
Researchers found that artificial intelligence tools can accurately predict disease for patients with typical symptoms but struggle with those exhibiting atypical symptoms. Human oversight is necessary for high-quality patient-centered care when using AI as an assistive tool.
Teaching theory of mind to robots to enhance collaboration
Researchers at Duke University have developed a new framework called HUMAC that enables robots to collaborate like humans by teaching them Theory of Mind. After just 40 minutes of guidance, robot teams exhibited strong collaborative behaviors and achieved high success rates in simulations and physical tests.
New global index defines what makes digital economies resilient and inclusive
The Digital Evolution Index reveals a crucial inflection point in the global digital landscape, with slowing growth and plateauing digital inclusion metrics. Emerging post-pandemic challenges include the 'winner-takes-most' scenario driven by AI, highlighting the need for resilient digital economies.
Online translators could easily learn Navajo, related languages, study suggests
A Dartmouth team's AI model recognized Navajo with near-perfect accuracy, identifying related languages such as Apache and Native Alaskan languages. The study suggests that this technology could be a bridge to including smaller languages in online translation services.
Researchers develop a novel vote-based model for more accurate hand-held object pose estimation
Researchers developed a novel vote-based model for accurate hand-held object pose estimation, addressing issues with existing approaches. The new framework achieves significant improvements in accuracy and robustness, enabling robots to handle complex objects and advancing AR technologies.
Obesity disrupts “reaction time” to starvation in mice
Researchers found that obesity causes a disruption in the liver's ability to adapt to starvation, specifically in the temporal coordination of molecules. This suggests that obesity makes the body more vulnerable to the negative effects of starvation, despite no significant structural disruptions in the molecular network.
‘Odd’ objects that adapt and move without a brain
Researchers develop active metamaterials that can autonomously roll, crawl, and wiggle over unpredictable terrain, including uphill and obstacles. These 'odd' objects achieve motion through unusual interactions between motorized building blocks, demonstrating decentralized and robust locomotion.
Machine learning maps animal feeding operations to improve sustainability
Researchers developed a machine learning model that predicts the presence of animal feeding operations with high accuracy, filling a data gap crucial for managing their environmental impacts. The model uses predictors such as surface temperature and phosphorus levels to identify locations without relying on aerial images.
What’s on the horizon for Australia’s road safety?
A new study forecasts Australia's road traffic fatalities to rise to 998 by 2030 and 715 by 2050, with older drivers and male motorcyclists at the greatest risk
Chicken ‘woody breast’ detection improved with advanced machine learning model
A new machine learning model, NAS-WD, has improved the accuracy of detecting 'woody breast' in chicken meat to 95%, allowing for better quality assurance and customer confidence. The model uses hyperspectral imaging to analyze complex data from images, enabling more accurate detection than traditional methods.
Human civilization at a critical junction between authoritarian collapse and superabundance
A new study warns that human civilization is poised for a significant transformation as industrial civilization declines, giving rise to a postmaterialist, clean energy-based system. Rising authoritarianism poses a threat to this transition.
Transportation institute awarded nearly $1 million in trucking education grants
The Virginia Tech Transportation Institute received nearly $1 million in grants to develop and enhance tractor-trailer educational programs. The programs focus on advanced driver assistance systems (ADAS) and safer driving outreach, aiming to save lives by educating drivers on the benefits of these technologies.
Study offers improvements to food quality computer predictions
A study from the University of Arkansas System Division of Agriculture has improved food quality computer predictions by using human perception data. The researchers trained a computer model to mimic human adaptation to environmental conditions, resulting in more consistent predictions under different lighting conditions.
Adaptive-optical 3D microscopy for microfluidic multiphase flows
Researchers developed a novel adaptive optics approach to correct dynamical aberrations in optical microscopy, enabling accurate three-dimensional flow measurements. The system reduces measurement uncertainty, paving the way to better understanding water droplet formation and detachment mechanisms for fuel cells.
Adaptive 3D printing system to pick and place bugs and other organisms
A first-of-its-kind adaptive 3D printing system developed by the University of Minnesota Twin Cities researchers can identify organism positions and safely move them to specific locations for assembly. This technology saves time and money in bioimaging, cybernetics, and cryopreservation.
UCF launches inaugural mentorship, scholarship initiative for students in AI
UCF's STRONG-AI initiative aims to uplift bright, low-income undergraduate students in pursuing well-rounded AI education through faculty and peer mentorship and scholarship. The program has received over 150 applications and will select 10-15 students annually based on financial aid eligibility and academic success.
From 'CyberSlug' to 'CyberOctopus': New AI explores, remembers, seeks novelty, overcomes obstacles
Scientists have developed an AI that can navigate new environments, seek rewards, map landmarks and overcome obstacles using a novel approach inspired by the brain circuits of sea slugs and octopuses. The new AI, called CyberOctopus, has the ability to explore and gather information while learning on the job.
Simplicity versus adaptability: Understanding the balance between habitual and goal-directed behaviors
A new study on learning has provided insights into the balance between habitual and goal-directed behaviors, with implications for AI development. The research suggests that a balance between these two types of behavior is necessary for efficient and adaptable decision-making in AI systems.
Deep learning reveals molecular secrets of explosive perchlorate salts
Researchers developed a novel deep learning method to study crystal structure and molecular interactions of perchlorate salts. The analysis revealed that the explosives' nature is linked to chemical bonding and intermolecular interactions.