Institute for Applied Systems Technology Bremen

A novel, temperature‐controlled post‐consolidation unit is developed to test its potential to improve the melt impregnation process used to manufacture continuous fiber‐reinforced filaments for additive manufacturing of high‐performance thermoplastics. Preliminary laboratory characterizations have identified Hexcel HexTow AS7 and Arkema Kepstan 6004 as suitable dry carbon‐fiber roving and thermoplastic matrix material, respectively. Filaments are processed by melt impregnation with varying parameters (e.g., different temperatures), as well as with and without the post‐consolidation unit. A thermographic camera is used to monitor the process and observe the temperature distribution of the filament during post‐consolidation. The filament quality is determined by a 3‐axis measurement device (filament diameter), micrographs, and X‐ray microscopy. The test results show that the temperature of the filament drops below the glass transition temperature over a production length of less than 100 mm when cooled by air. Using the post‐consolidation unit keeps the filament above the glass transition temperature of the matrix (160 °C) for a production length of 550 mm, allowing compaction by the integrated grooved rollers. The test results show that post‐consolidation reduces defects, improves the roundness of the filament cross‐section, and minimizes diameter variation.

The increasing presence of polluting chemicals in man-made wastewater poses significant environmental and health risks. In this work, the potential for removal of selected pollutants is discussed with results of molecular dynamics simulations.

The meltblow process was utilized to produce nonwoven fabrics from polylactic acid (PLA) in order to evaluate the potential of this bio-based polyester as an alternative to conventional polyolefins. Therefore, it was opposed against polypropylene (PP) meltblown as benchmark material, which is commonly used as a filter layer, for example in particle-filtering face masks. PLA was found to be processable at more moderate thermal process conditions, while its fiber diameters showed higher sensitivity to process parameter variation. This allowed the controlled deposition of both fine-fiber meltblown webs, but also coarse – spunbond-like – meltblown fiber webs. These showed minor mechanical performance but sufficient haptics and equal air permeability to spunbond PLA. Electret treatment of fine fiber PLA meltblown (1.43 µm average diameter) resulted in a raise of particle filtration performance (in paraffin oil) from 70% to >97%, exceeding the performance of PP meltblown of same grammage and fiber diameter distribution. The effect was also traceable over storage time (90 days), with just a slight reduction of 5% over the first days. These findings highlight PLA’s potential as a high-performance, bio-based alternative for filtration applications. Furthermore, the results contribute to the understanding of bio-polymer processing by comparing the processability, fiber morphology, and resulting performance of PLA and PP under similar process conditions.

Abstract Herein, two Cu‐based model electrocatalysts, Cu nanosheets modified with Cd single atoms (Cd‐SACs@Cu NS) and unmodified Cu nanosheets (Cu NS), were rationally designed to investigate the C–C coupling to C 2+ alcohols mechanisms in CORR. Compared to Cu NS, Cd‐SACs@Cu NS exhibits a significantly enhanced Faradaic efficiency (FE) for C 2+ alcohols of nearly 70% at an unprecedented current density of 1100 mA cm −2 and maintains a higher FE (>50%) over a broad range of current densities (100–2100 mA cm −2 ) under alkaline conditions in a flow cell. In situ characterizations and theoretical studies reveal that the enhanced selectivity toward C 2+ alcohols on Cd‐SACs@Cu NS is attributed to the improved H 2 O dissociation at atomically dispersed Cd sites and enhanced CO adsorption on the paired Cu atoms adjacent to single Cd atoms. These effects synergistically facilitate the spillover of hydrogen atoms from Cd to Cu sites, thereby promoting the protonation at the β‐carbon of *CH 2 CHO, which leads to the selective formation of C 2+ alcohols. In contrast, the unmodified Cu NS is prone to promoting the cleavage of the C─O bond, thereby facilitating the generation of C 2 H 4 .

Automatic license plate detection and recognition systems are crucial in modern intelligent transportation systems, including traffic control, toll collection, and parking management. This paper presents a real-time, AI-driven, Azure IoT-based automatic license plate recognition system tailored for Indian traffic environments. The system utilizes an Intel RealSense D435i depth camera paired with an Nvidia Jetson Xavier NX to capture vehicle images and extract license plate numbers using a lightweight deep learning pipeline: YOLOv8 for detection and EasyOCR for robust optical character recognition. The recognized license plate text is enriched with location and timestamp metadata and transmitted to the cloud through the Azure IoT Edge runtime. At the cloud level, the Azure IoT Hub enables seamless ingestion and routing of telemetry data to various services, including Azure Stream Analytics, SQL Database, and App Services. This infrastructure supports AIbased analytics, real-time alerting, and long-term data storage. A Web and mobile application interface - powered by Azure APIs and Power BI dashboards - allows traffic authorities to visualize vehicle data, monitor violations, and receive alerts for blocked or duplicated plates. The system achieves a detection accuracy of 95.4 % and an OCR precision of 92.3 % while maintaining an average throughput of 15 FPS. This scalable, low-latency automatic license plate recognition system offers an effective solution for intelligent traffic management and law enforcement in smart city applications. The proposed system offers a scalable and adaptable solution for automated license plate recognition in various applications.

Patients often want to be involved in decisions about their health. However, to actively participate in the decision-making process and make informed choices, they often require support. Decision coaching can be offered as a non-directive form of decision support provided by trained health care professionals. So far, decision coaching has only been tested in pilot projects in Germany. To investigate the determinants of practice that influence implementation, decision coaches were interviewed in a qualitative study about their practical experiences with decision coaching and regarding facilitators and barriers in its implementation. Guided expert interviews were conducted from June 2023 to November 2023 with decision coaches from model projects in Germany. The decision coaches were asked about their tasks, work processes, framework conditions, understanding of their role, self-perception, the decision support tools used and barriers and facilitators for the implementation of decision coaching. Sociodemographic data were collected anonymously. The verbatim transcribed interviews were analysed using content analysis according to Kuckartz with MAXQDA. Barriers and facilitators of the implementation of decision coaching were analysed using the Consolidated Framework for Implementation Research (CFIR 2.0) framework. Seven guided interviews were conducted with experienced female decision coaches (Ø 56 years) from pilot projects in oncology and neurology. While some elements were sustained in practice, only one project continued long-term. Overall, the decision coaches strongly identified with their role and patient acceptance was high. Thematic analysis revealed seven main categories with 32 subcategories. Crucial factors influencing the implementation included structured training, flexible delivery formats, institutional support, dedicated space and funding. Although decision coaching was perceived to enhance patient preparedness and physician-patient communication, structural challenges remain. The coaches perceived their role as supportive in patients’ decision-making. However, the effectiveness and implementation of decision coaching are strongly influenced by several contextual factors that could guide the refinement of interventions and the selection of implementation strategies. Not applicable.

Abstract Hydrogen offers a high-energy, carbon-free fuel alternative; however, conventional flame-based hydrogen combustion poses challenges, including NO x emissions and the risk of flame flashback. Catalytic combustion provides a safer, low-temperature approach to hydrogen utilization, but realizing it within compact, integrated systems have remained a significant challenge. This study introduces an innovative approach to hydrogen catalytic combustion by directly integrating noble metal single quantum-crystallites of Pt and Ru within a porous silica aerogel matrix embedded in a silicon chip. This configuration enables deep nanoparticle ( np ) penetration throughout the aerogel network, maximizing the catalytic surface area and providing efficient on-chip hydrogen combustion. The np @aerogel systems are systematically synthesized and incorporated within silicon chips equipped with a polyimide membrane and Pt thermal structures. This unique setup allows for direct, real-time characterization of hydrogen catalytic combustion by measuring resistance changes in an embedded thermistor. The Pt@SiO₂ system demonstrates a rapid and substantial temperature increase of up to 40 K upon hydrogen exposure, independent of both preheating and Pt concentration, underscoring its robustness and adaptability for micro-scale hydrogen combustion. This on-chip integration of np @aerogel catalysts marks a significant advancement for hydrogen-based energy applications, offering a compact, scalable platform for efficient catalytic combustion. This approach opens pathways for applications in thermoelectric generators and other micro-reactor technologies where controlled, localized energy generation is critical.

In typical power electronics applications, semiconductor switches are subject to various stress conditions during operation. Depending on the load, temperature changes occur within the chip, causing thermomechanical stress on the interconnections within the package. One possible failure mechanism induced by cycling loads is bond wire lift-off. In this paper, bond wire lift-off is treated as if the wire were completely removed from the package. Based on this assumption, the study investigates how parasitic components increase and how this affects the switching behaviour of both single and parallel SiC-MOSFETs. To accurately measure the current through the devices an M-Shunt is integrated into the source path as a low-inductive current sensor. Furthermore, the paper introduces the use of the M-Shunt as a sensor for measuring Kelvin-source currents. It examines how the removal of source bond wires affects the transient and steady-state Kelvin-source currents. By means of artificial intelligence, the change in the Kelvin-source currents might be used as a criterion to assess the health condition and remaining lifetime of the SiC-MOSFET.

This double-blind cluster-randomized controlled study examined the effectiveness of a multilingual physical-digital recruitment strategy to increase the involvement of linguistically diverse households in the German sub-survey of the World Health Organization European Childhood Obesity Surveillance Initiative. Students in primary schools received German study invitations either alone or with a supplemental flyer containing a QR-code for accessing translations online. A total of 114 classrooms (2306 students) received the flyer, while 115 classrooms (2276 students) served as control group. Main outcomes were active of responses to invitations and study participation. Response proportions were 48% overall, 48.6% in the flyer and 47.4% in the control group. Participation proportions were 26.7% overall, 27.8% in the flyer and 25.4% (571/2251) in the control groups. The flyer increased the odds of participation (14.9, 95% CI 1.003-1.315), but not the odds of active responses. In medium-social-level schools, response and participation proportions in the flyer group rose by 1.5% (ROR 1.015, 95% CI 1.003-1.027) and by 2% (ROR 1.020, 95% CI 1.005-1.034), respectively, for each percentage point increase in the proportion of non-German speaking households per classroom. The flyer's effect in low- and high-social level schools was inconclusive. Although modest in magnitude, the addition of the flyer increased overall participation and, in particular, the involvement of linguistically diverse households in medium-social-level schools.