University of Applied Sciences Technikum Wien

Abstract Ensuring food security goals through the provision of sustainable energy and food without compromising environmental sustainability is the current requirement amid threats from climate change, uncontrolled global population, and scarcity of natural resources. Increased agricultural dependence on chemical inputs to boost food productivity has become less ecologically efficient, threatening ecological balance. The utilization of favorable soil microbes as a key substitute for chemical fertilizers can prove a potential solution to this riddle. Cyanobacteria, a group of photosynthetic prokaryotes, have been considered as potential sources of various bioenergy forms (biogas, biodiesel, and bioethanol) due to their native abilities to efficiently harvest solar energy and convert it into biomass by simple dependence on CO 2 , nutrients, and water. Efficient application of cyanobacteria can mitigate climate change by lowering the greenhouse gas emissions and bioremediation of contaminated soil and water. This article reviews specific knowledge of the ecological mechanisms, including carbon sequestration, nitrogen fixation efficiency, and soil microbial enhancement, about how these mechanisms can be further improved through the incorporation of cyanobacteria. This suggests that future research should focus on utilizing already available indigenous cyanobacterial strains and searching for new species to explore the symbiotic mechanisms executed with other beneficial soil microbes, from an environmentally friendly perspective, toward the management of the rhizosphere to encourage plant growth and development. Moreover, it is necessary to explore the composition, diversification, distribution, and environmental factors of cyanobacteria in rice ( Oryza sativa L.)‐based soils under different ecological conditions to better utilize the biotechnology tool.

Deploying deep neural networks (DNNs) in real-time, resource-constrained edge environments remains a significant challenge due to the increasing depth and computational demand of state-of-the-art models. While state-of-the-art split computing and early-exit architectures independently offer solutions to reduce inference latency and offloading overhead, they are rarely integrated in a unified framework. This paper introduces <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BranchySplit</i>, a novel architecture that combines dynamic state-of-the-art split computing with early exit mechanisms to accelerate inference in edge-cloud systems. Our method employs a shared dynamic autoencoder to support multiple partitioning points, enabling runtime adaptability based on varying system conditions. By integrating early exit branches at intermediate layers, the model can terminate inference early when confidence is sufficient, further reducing computation and communication. Extensive experiments on Caltech-256 and Tiny-ImageNet dataset using four backbone models (VGG-16, ResNet-50, Efficientformerv2-S2, and ConvNeXT-small) show that our approach consistently outperforms the traditional split computing approaches, achieving up to 2.7<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</i> latency speed-up compared to an all-in-server solution, while maintaining comparable or even improved accuracy. In addition, our proposed dynamic compression pipeline achieves a higher speed-up compared to the pipeline in similar work, i.e. SPINN, with up to 10<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</i> for ResNet50 on caltech-256. These results demonstrate the efficacy of BranchySplit in enabling low-latency, resource-efficient vision inference in heterogeneous edge-cloud settings.

Multiple programming tasks require synchronous collaboration between developers, giving rise to real-time collaborative programming tools that enable simultaneous editing of shared source code. However, most existing tools operate at the text level, propagating every keystroke–including syntactically invalid ones–without considering program structure. This results in excessive communication overhead, frequent propagation of build-breaking states, and poor synchronization. A major consequence is noticeable lag, especially under unstable network conditions, as collaborators are overwhelmed with unnecessary updates that disrupt their workflow and degrade the shared coding experience. In this paper, we introduce a novel structure-aware propagation model that transmits only syntactically valid code changes. For evaluation we implemented our tool as an IntelliJ plugin and evaluate it against three industry-standard tools–VS Code Live Share, Code With Me, and Replit–across eight representative programming scenarios. Our results show that it significantly lowers the number and size of propagated messages while maintaining consistent, buildable program states. Our findings demonstrate the potential of structure-aware propagation as a foundation for the next generation of real-time collaborative programming environments.

Visual servoing enables robots to precisely position their end-effector relative to a target object. While classical methods rely on hand-crafted features and thus are universally applicable without task-specific training, they often struggle with occlusions and environmental variations, whereas learning-based approaches improve robustness but typically require extensive training. We present a visual servoing approach that leverages pretrained vision transformers for semantic feature extraction, combining the advantages of both paradigms while also being able to generalize beyond the provided sample. Our approach achieves full convergence in unperturbed scenarios and surpasses classical image-based visual servoing by up to 31.2% relative improvement in perturbed scenarios. Even the convergence rates of learning-based methods are matched despite requiring no task-or object-specific training. Real-world evaluations confirm robust performance in end-effector positioning, industrial box manipulation, and grasping of unseen objects using only a reference from the same category. Our code and simulation environment are available at: https://alessandroscherl.github.io/ViT-VS/

Abstract Rapid and reliable DNA extraction from urine is a critical bottleneck in advancing molecular diagnostics for urinary tract infections (UTIs) in both centralized and decentralized settings. Here, we present an ionic liquid-based DNA extraction method (IL-DEx) that enables recovery of bacterial DNA from urine samples in under 30 minutes using minimal equipment and no hazardous chemicals. IL-DEx was benchmarked against a widely used commercial kit (QIAamp DNA Mini Kit, QIAGEN) using reference strains, clinical isolates, and spiked urine samples. For gram-negative bacteria, IL-DEx achieved comparable DNA yields (47–102% relative efficiency), while recoveries from gram-positive bacteria were lower (0.7–8%) but sufficient for downstream detection. Quantitative PCR (qPCR) revealed linear DNA recovery across five to six orders of magnitude (10 8 –10 2 CFU/ml, R 2 &gt;0.99), with detection limits of ∼10 2 –10 3 CFU/ml for gram-negatives and ∼10 3 –10 4 CFU/ml for gram-positives using 1 ml urine. Clinical evaluation with 13 patient urine samples (ten culture-positive, three culture-negative) demonstrated that IL-DEx reliably enabled pathogen detection by qPCR and full-length 16S rRNA gene sequencing (Oxford Nanopore). Performance was comparable to three other extraction methods tested head-to-head, including the QIAamp DNA Mini Kit (QIAGEN), the MagaZorb DNA Mini-Prep Kit (Promega), and a phenol-chloroform extraction method. These findings establish IL-DEx as the first ionic liquid-based approach evaluated for DNA recovery from clinical urine samples, providing a fast, simple, and low-cost method suitable for integration into molecular workflows for UTI diagnostics across diverse laboratory and clinical settings. Importance Urinary tract infections (UTIs) are among the most common infections worldwide and a major driver of antibiotic use. Rapid and accurate diagnosis is critical to guide therapy, reduce inappropriate antibiotic prescriptions, and improve patient outcomes. While molecular diagnostics can drastically reduce time to identify uropathogens, their implementation remains constrained by upstream DNA extraction – a step that is often laborious, cost-intensive, or incompatible with rapid diagnostic workflows. We developed a fast, simple, and low-cost DNA extraction method (IL-DEx) that uses an ionic liquid and magnetic beads to recover bacterial DNA directly from urine. IL-DEx eliminates hazardous reagents and complex equipment while delivering performance comparable to established extraction kits. By streamlining this critical pre-analytical step, IL-DEx enables faster molecular diagnostics and broadens access to modern UTI testing. Its simplicity and robustness position it as a valuable tool for improving diagnostic speed, antimicrobial stewardship, and patient care across healthcare settings.