Tech & AI

Perovskitbaserade lysdioder når nu effektivitet som gör kommersiell konkurrens möjlig. Forskare vid NanjingTech, Zhejiang University och Linköpings universitet presenterar en lösningsbearbetad perovskitLED med självorganiserade kvantbrunnar som uppnår extern kvanteffektivitet på 11,7 procent och energiomvandlingseffektivitet på 5,5 procent vid 100 mA/cm². Konstruktionen använder bandgapskonfinering — högre energiegap omsluter lågare områden som genererar ljus — för att minska ej-strålande rekombination från defekter och strömlekage. Det kritiska genomslaget är att gränssnitten mellan bandgapen inte orsakar luminescensslöckning som tidigare förutsagts. För leverantörer och produktchefer betyder detta att perovskiter kan bli konkurrenskraftiga mot OLED och konventionella halvledarlysdioder inom displaytillämpningar. Stabilitet och högeffektivitet öppnar vägen för praktiska applikationer inom några år.

Forskare vid Högskolan i Borås har utvecklat en prediktionsmodell för att beräkna den specifika kemiska exergin i kommunalt avfall — en parameter som är avgörande för att värdera avfallsomvandlingen till energi. Modellen bygger på elementarsammansättningen (kol, väte, syre, kväve, svavel, klor) och använder statistisk analys av 56 avfallsfraktioner samt validering på 30 ytterligare prover. Forskarna analyserade även entropi för 117 organiska ämnen och bedömde bidraget från oorganisk materia. Resultaten visar att värmevärdet dominerar över entropins inverkan när man skattar avfallsexergin. Modellen korrelerar väl med uppmätta värden och är jämförbar med tidigare modeller. För energi- och avfallshanteringsbranschen möjliggör detta mer exakt dimensionering av förbränningsanläggningar och värmeåtervinningssystem, vilket förbättrar ekonomin i avfallsvärmekonvertering och bidrar till effektivare resursåtervinning.

A new IEEE special issue explores how lightweight machine learning can run directly on edge devices in future 6G networks, reducing latency and infrastructure costs. The approach could reshape how telecom companies deploy AI-driven services without relying on cloud processing.EN

Researchers ran an open competition where teams worldwide tested computational methods on real coronavirus drug data, benchmarking which approaches best predict drug potency and safety. The results could reshape how pharma companies prioritize AI tools and accelerate discovery timelines by validating—or eliminating—unproven methods before investing millions.EN

Researchers have developed lightweight monitoring systems that finally let network operators see the actual latency their customers experience, rather than relying on incomplete measurements. The breakthrough matters because latency directly impacts user experience and revenue for ISPs and content providers—and they've proven it works in real production networks.EN

Researchers have identified a critical gap between laboratory successes in "superlubricity"—nearly frictionless lubrication—and practical use in industrial machinery like gears and bearings. The findings suggest that decades of promising lab results may not translate to factories without rethinking how these lubricants interact with actual machine components and operating conditions.EN

Researchers developed a computational method to forecast how polymer materials organize in organic solar cells, a critical factor for device efficiency. The advance could accelerate solar panel design by replacing costly trial-and-error experiments with physics-based simulation.EN

Researchers have identified a class of tracking problems where a supposedly superior filtering algorithm produces no better results than a basic alternative—but engineers have no easy way to detect it beforehand. The finding matters because companies and agencies relying on these systems for autonomous vehicles, robotics, and surveillance may be wasting computational resources without realizing it.EN

Researchers have developed Criminator, an extended reality platform that lets law enforcement officers—even those without 3D animation skills—quickly recreate and test theories about how crimes unfolded. The tool could reshape how evidence is presented in court and speed up investigations, though legal questions remain about whether animated reconstructions should count as admissible evidence.EN

Researchers in São Paulo tested satellite-based radar technology to track ground movement during subway construction, achieving millimeter-level precision at a fraction of traditional sensor costs. The approach could reduce construction disputes and improve safety monitoring across large urban infrastructure projects worth billions of dollars.EN

A comprehensive review of 384 studies shows construction companies understand circular economy principles—reusing materials, reducing waste, cutting costs—yet rarely implement them at scale. The gap between knowledge and practice signals a major market opportunity for suppliers and contractors who can bridge the implementation divide.EN

A new systematic review reveals that virtual reality networks are engineered to handle individual users, not the complex demands of groups collaborating remotely. As companies invest billions in VR for training, healthcare, and social platforms, this mismatch is degrading shared experiences—and researchers say fixing it requires rethinking how networks prioritize data flow across multiple connected users.EN

Researchers trained neural networks on nearly 2 million eye-region images to develop periocular recognition—a biometric authentication method that works from photos of just the area around the eyes. The advance could reshape identity verification for banking, border security, and mobile devices, especially in low-light or masked scenarios where full-face recognition fails.EN

Researchers have solved a long-standing inconsistency in how scientists set boundary conditions for compressible fluid flow equations—a fundamental problem underlying everything from aircraft design to weather forecasting. The fix uses a new matrix-based analysis method that ensures complex nonlinear models behave consistently with simpler linear versions, potentially accelerating computational modeling across aerospace, energy, and climate sectors.EN

Researchers have developed a lightweight artificial intelligence system that detects and locates pipeline leaks in real time using acoustic signals, while running on low-power edge devices. The advance could save oil and gas operators billions in prevention costs and environmental liability by catching failures early on offshore platforms.EN

A new review reveals that metal components made via 3D printing are surprisingly vulnerable to stress corrosion cracking—a hidden failure mode that can destroy parts in corrosive environments like oil rigs or aircraft. The findings matter because manufacturers betting on additive manufacturing for complex parts need better design rules and testing protocols before deploying these components in high-stakes applications.EN

Researchers have shown that optimizing neural network algorithms—rather than building better hardware—can dramatically improve how well electronic nose sensors detect dangerous gases and spoiled food. For manufacturers, the finding suggests a faster, cheaper path to wider adoption of these safety-monitoring devices across industries from food production to chemical plants.EN

Researchers have created the first unified model for how autonomous systems—from self-driving cars to industrial robots—should represent and manage different types of uncertainty. The work addresses a critical gap in AI reliability: existing approaches use inconsistent terminology and fail to distinguish between software, hardware, and autonomous decision failures, making it harder for companies to build trustworthy systems.EN

Researchers have developed a communication method that dramatically reduces the amount of data needed to transmit information while guaranteeing privacy protection. The technique could reshape how companies handle sensitive data in cloud systems, financial networks, and IoT devices—cutting bandwidth costs and security risks simultaneously.EN

Researchers have developed a computational framework that finally makes it practical to accurately simulate pulsating heat pipes—passive cooling devices critical for high-performance electronics. The work resolves longstanding modeling uncertainties that have stalled adoption of these compact, efficient cooling solutions in data centers and advanced semiconductors.EN

Researchers have developed a faster statistical method for analyzing huge amounts of data without losing accuracy. The breakthrough matters for companies in finance, healthcare, and telecommunications that need to extract insights from billions of data points quickly—potentially cutting computational costs while maintaining reliable results.EN

Researchers built a tabletop system where small robot carts physically move to represent data points, helping people explore information faster and remember spatial patterns better than touchscreens. The finding suggests physical, tangible interfaces could improve how organizations visualize and interact with complex datasets—potentially reshaping workplace tools for analytics and decision-making.EN

Researchers developed an AI control system that automatically switches between heat pumps and traditional district heating to minimize energy costs in real time. The technology could slash utility bills for millions of European homes already connected to district heating networks while reducing grid strain during peak demand periods.EN

A new evaluation of low-cost volatile organic compound sensors reveals significant accuracy problems that could undermine building ventilation systems designed to cut energy costs. The sensors showed wildly inconsistent readings—sometimes off by a factor of five—and identical units produced different outputs, raising questions about their real-world deployment in offices and homes.EN

Researchers have perfected a micro-scale laser coating process that deposits ultra-thin, defect-free protective layers on nuclear fuel rods with minimal heat damage. The breakthrough could accelerate deployment of advanced nuclear fuels while reducing manufacturing costs and waste—critical for the nuclear industry's expansion plans.EN