SK hynix is exhibiting its research and development capabilities and innovation roadmap at SEMICON Korea 2026, from February 11-13, reinforcing its leadership in the global memory market. This premier semiconductor industry event, organized by Semiconductor Equipment and Materials International (SEMI), features participation from 550 companies across 2,411 booths, showcasing advanced technologies and facilitating discussions among industry professionals on emerging trends.

Keynote speakers from major global semiconductor firms address innovative technologies and optimization strategies relevant to the AI era, emphasizing supply chain development and collaboration. Among the highlights, Lee Sunghoon, Senior Vice President of R&D Process at SK hynix, discussed the company’s AI-driven advancements in a keynote titled “The Inflection Point Has Arrived: Driving Innovation Towards the Future of Memory Technology.” He elaborated on the escalating complexity in semiconductor manufacturing processes, particularly in DRAM and NAND technologies.

AI applications showcased during the event include process simulations developed in collaboration with NVIDIA, enabling faster evaluations of new materials and optimal processing conditions. This approach minimizes the number of experiments needed, significantly enhancing time efficiency in R&D compared to traditional methods. Lee underscored the importance of AI models and data management in driving innovation, suggesting a shift toward AI-based R&D could tackle industry-wide challenges.

At the SEMI Technology Symposium, SK hynix engineers provided insights into the latest technologies, covering diverse topics such as extreme ultraviolet lithography and advanced packaging for AI on-device applications. The event ultimately emphasizes SK hynix's commitment to maintaining its technological edge and fostering innovation in the AI memory sector.

Lightweight, high-strength polymer composites are crucial in modern engineering. However, their manufacturing processes are often complicated and slow, typically relying on manual adjustments. A recent analysis in Frontiers of Chemical Science and Engineering outlines a groundbreaking approach that leverages artificial intelligence (AI) to create fully integrated, self-optimizing production systems.

This study showcases how AI can effectively connect design, process selection, and quality control, moving beyond traditional methods that consider manufacturing steps in isolation. By employing machine learning and digital twins, the proposed framework optimizes the entire production chain cohesively.

Dr. Zijie Wu from Yaoshan Laboratory emphasizes that the integration of AI with composite material design shifts the industry from experience-based practices to data-driven strategies. AI enhances our understanding of material behavior, process parameters, and final performance, enabling the production of lighter, stronger, and more reliable components while minimizing waste.

One significant application is the use of physics-informed neural networks to model the curing stage. These AI models analyze historical sensor data to optimize heating and pressure curves, leading to cycle time reductions of up to 30% and lower energy consumption. Another innovative integration combines hot pressing with injection molding in a single step, streamlining the manufacturing process for structural bases and functional features.

The technology directly addresses industry challenges such as high costs of prototyping, inconsistent part quality, and scaling new materials. Companies like Boeing and Airbus are already testing AI tools for autoclave optimization and automated fiber placement, achieving gains in precision and throughput.

Moreover, the implications for sustainability are profound. Smarter manufacturing of composites not only enhances their lightweight benefits in transport but also further minimizes resource usage. This research presents a clear pathway to modernizing composite manufacturing through AI, improving product quality and agility while supporting a more sustainable industrial future.

Draslovka a.s., a frontrunner in sustainable chemical technologies and AI services for mining, has forged a strategic alliance with Avathon, a leader in Operations Autonomy. This partnership aims to rapidly integrate intelligent, autonomous, data-driven operations within the global mining sector.

The collaboration leverages cutting-edge technologies from both companies to enhance mining operations significantly. Mining firms can expect increased mineral recovery rates and performance stability through real-time AI-enabled mineralogy and metallurgical insights. The alliance also aims to cut down reagent consumption, energy use, and operational variability, while automating maintenance and supply chain processes, thus improving asset performance and minimizing unexpected downtime.

Central to this initiative is Draslovka’s offering of Blue Cube™ online mineral analyzers, which supply critical real-time process data necessary for optimizing decision-making. Alongside this, MetOptima™ provides an AI-driven optimization engine that analyzes live data to identify inefficiencies and recommend optimal control strategies. Pavel Bruzek, CEO of Draslovka, emphasized that this integration will empower mining operations with real-time optimization and autonomous decision capabilities.

Avathon’s Autonomy Platform enhances this synergy by applying industrial AI to streamline operational workflows across diverse sites. It provides a unified environment for managing maintenance, health, safety, and supply chain activities. As Avathon CEO Pervinder Johar noted, this collaboration bridges the gap between plant-level optimization and enterprise-level operational resilience.

Overall, the Draslovka-Avathon alliance promises to revolutionize the mining industry by fostering operational efficiency, sustainability, and real-time decision-making capabilities, ultimately driving significant advancements in mine management.

Detecting concealed explosives and chemical threats is a vital aspect of global security, but existing technologies often have operational limitations. Traditional X-ray scanners and millimeter-wave imaging can detect suspicious shapes but lack chemical specificity. On the other hand, precise chemical sensors, like mass spectrometry or trained dogs, require close proximity to the target, posing safety risks, especially in crowded or volatile environments. Terahertz spectroscopy has emerged as a potential solution, capable of penetrating materials like clothing and plastic without ionization damage. However, conventional methods struggle under real-world conditions, where chemical signatures are often obscured.

Researchers from UCLA, led by Professors Mona Jarrahi and Aydogan Ozcan, have developed an advanced chemical imaging system that combines high-performance terahertz time-domain spectroscopy with deep learning. This innovative system can detect and classify explosives, even when concealed or irregularly shaped. By using plasmonic nanoantenna arrays for terahertz generation and detection, the system achieves a wide dynamic range and bandwidth. Instead of relying on averaged spectra, it analyzes individual time-domain pulses reflected from the sample.

The deep learning framework applied here significantly enhances the system's efficacy, discerning unique chemical signatures from environmental noise. Demonstrated through rigorous blind tests on various chemical compounds, the system achieved an impressive 99.42% average classification accuracy for exposed samples and maintained 88.83% accuracy for explosives concealed under opaque coverings. This technology holds transformative potential for rapid, stand-off chemical imaging, benefiting applications in security screening, pharmaceutical manufacturing, and industrial quality control.