GAIKER Technology Centre

Current literature on health hazards associated with micro- and nanoplastics (MNPs) is largely influenced by studies that insufficiently account for potential microbial contamination of their test materials. This may lead to misinterpretation of outcomes, as the test materials may be incorrectly considered pristine MNPs. The present study screened eight MNP test materials for microbial contaminants using Toll-like receptor (TLR) reporter cells for TLR2 and TLR4 and the commonly used Limulus amebocyte lysate (LAL) assay. Our results show that MNPs testing negative for endotoxins, based on the absence of TLR4 activation and negative LAL results, may still contain microbial ligands that selectively activate TLR2. Moreover, five of the eight MNP test materials contained microbial ligands capable of activating TLR2 and/or TLR4. Compared to the LAL assay, TLR4-based screening effectively detected endotoxin contamination. Overall, we found that the TLR reporter cell assay provides broader coverage than the LAL assay in detecting microbial ligands, which appear to be highly prevalent in MNP test materials.

Large Language Models are increasingly used for high-level robotic reasoning, yet their latency and stochasticity complicate their direct use in low-level control. Moreover, extracting actionable navigation cues from multimodal context incurs inference costs that are challenging for embedded platforms. We present a plug-and-play framework that augments a finite-state machine with asynchronous velocity suggestions generated by a Large Language Model, using an off-the-shelf DistilGPT-2 model running on-device on a Jetson AGX Orin. The system extracts task-relevant cues from the current context and integrates them only if they satisfy deadline, schema, and kinematic validation, thereby preserving a deterministic 50 Hz control loop with a <5 ms fallback path. We compare multiple Large Language Models for embedded robot control and quantify trade-offs among model size, inference time, and output validity. To assess whether the Large Language Models add value beyond signal processing, we include an ablation against a standard smoothing baseline; the results indicate that the Large Language Models contribute anticipatory, context-dependent adjustments that are not captured by filtering alone. Experiments in Gazebo and on a real TurtleBot3 reduce the final position error from 0.246 m to 0.159 m and improve trajectory efficiency from 0.821 to 0.901 without increasing control-loop latency. Approximately 80% of the Large Language Models’ outputs pass validation and are applied. Overall, the framework reduces developer effort by enabling behavioral changes at the prompt level while maintaining interpretable, robust edge-based navigation.

This study investigates the microstructural evolution and wear behaviour of aluminium 7075-based metal matrix composites (MMCs) reinforced with high-entropy alloy (HEA) particles and fabricated via friction stir processing (FSP). A detailed characterisation of the grain refinement in the 7075 matrix was conducted, revealing significant dynamic recrystallization and grain size reduction induced by the severe plastic deformation inherent to FSP. The interaction between the matrix and HEA particles was analysed, showing strong interfacial bonding, which was further influenced by post-processing heat treatments. These microstructural modifications were correlated with the wear performance of the composites, demonstrating enhanced resistance due to the synergistic effect of precipitates and particle reinforcement. The findings highlight the potential of FSP as a viable route for tailoring surface properties in advanced MMCs for demanding tribological applications.

The Safe and Sustainable-by-Design (SSbD) framework by the EC-JRC (European Commission - Joint Research Centre) provides a structured approach to integrate safety and sustainability considerations from the earliest stages of chemical and material innovation. However, applying SSbD principles to advanced materials poses specific challenges due to their complex and diverse physicochemical properties. This work analyzes and maps hazard, exposure, fate and risk assessment methods and tools applicable to Steps 1, 2 and 3 of the EC-JRC SSbD framework, categorizing them across its three tiers to address different stages of product development. The analysis highlights the challenges of adapting conventional testing and modelling approaches to advanced materials, particularly for hazard assessment, and considers the relevance and limitations of tools originally developed for exposure and risk assessment of engineered nanomaterials when applied to broader advanced materials categories. An assessment of operational status, access conditions, and tool formats provides practical insights for researchers and industry stakeholders. The study identifies key methodological gaps and offers recommendations to improve and expand the current tool landscape. By providing a structured mapping of available resources and challenges, this work supports the effective implementation of SSbD principles, promoting the safe and sustainable development of advanced materials.

Pulmonary infections caused by <italic>Pseudomonas aeruginosa</italic> constitute a therapeutic challenge due to the ongoing difficulties in developing both effective and safe treatments. Inhaled tobramycin (TB) is a reference therapy, but the efficacy may be limited by its rapid clearance and limited penetration into the pulmonary mucus. Therefore, the development of advanced drug delivery systems is critical to enhance retention and bioavailability in the lung. The objective of this study was to evaluate the therapeutic capacity of TB encapsulated in liposomes without PEG (LP-TB) and with PEG (LPPEG-TB) compared to free TB to treat <italic>P. aeruginosa</italic> infection in <italic>in vitro</italic> lung models. The therapeutic efficacy of the compounds was evaluated at different time points (24 h, 48 h and 72 h) using a cellular lung model at air-liquid interface through Vitrocell Cloud 12 exposure system (VITROCELL^® Systems GmbH) and artificial pulmonary mucus (APM), with both models infected with <italic>P. aeruginosa</italic>. After testing the safety and minimum inhibitory concentration (MIC) of the compounds, it was observed that LP-TB and free TB were rapidly internalised by lung cells, while LPPEG-TB remained on the cell membrane for a longer time before internalisation. Furthermore, in the APM model, LPPEG-TB demonstrated greater ability to migrate through the mucus layer compared to LP-TB and free TB. These results suggest that LPPEG-TB are promising candidates for treating <italic>P. aeruginosa</italic> pulmonary infections, as their increased ability to penetrate mucus reduces mucociliary clearance and their prolonged retention in the cell membrane before cellular uptake allows a more sustained antibacterial activity.

Artificial Intelligence (AI) has traditionally prioritized high performance over compliance with functional safety standards such as IEC 61508. However, when AI systems are used in safety-related functions, it is essential to demonstrate that errors will not lead to malfunctions. This involves preventing systematic design-time errors and detecting and controlling runtime faults, as specified in IEC 61508. Moreover, ISO/PAS 8800 requires analyzing AI-specific development tools to identify and mitigate potential risks. In this paper, we take a step toward a safe end-to-end AI framework by focusing on systematic error avoidance in the implementation of You Only Look Once (YOLO), a widely used object detection model. A C-based version of YOLO-built on the Darknet framework-is analyzed using the Polyspace static analysis tool to assess MISRA C compliance. We apply corrective actions to eliminate violations, producing a MISRA $\mathbf{C}$-compliant implementation. In addition, we propose a runtime error detection mechanism using dual execution on a diverse platform and validate behavioral consistency using the COCO dataset. This approach supports the development of trustworthy AI systems by addressing both systematic errors and runtime detection.

Low Voltage Ride Through (LVRT) behaviour of Inverter Based Resources (IBR) during asymmetrical grid faults is dictated by the requirements of Grid Code (GC) at the Point of Common Coupling (PCC) of the renewable energy plants. Earlier GCs demanded only positive sequence current injection, which gave freedom to inverter manufacturers to implement different negative sequence injection strategies, while newer GCs define specific negative sequence current injection requirements. The existence of different negative sequence current control strategies during grid faults creates problems for line protection relays, as fault detection algorithms malfunction for certain angle relationships between the negative sequence voltage and current phasors. This paper analyses different negative sequence injection algorithms of IBRs and demonstrates certain relationships between sequence impedances of the inverter that can be used for improving line protection algorithms. The analytical findings are confirmed via Real Time simulations of a Type 4 wind farm connected to a power system with different network strength characteristics.

In this work, continuous ferromagnetic microwires were integrated into glass fiber reinforced composite (GFRP) plates. The microwires were placed in a single orientation maintaining a constant distance between them. A vector network analyzer (VNA) and two antennas were placed in the vicinity of the composite plates that was mounted in a universal testing machine. Therewith, reflection and transmission scattering parameters were determined by means of free space measurements, while simultaneously, the composite test specimens were submitted to tensile stresses. Force was progressively increased, and dynamic stress cycles were also applied. No external magnetic fields were applied to the composite structures during the measurements. The obtained results were compared with those generated with the same procedure applied on composite specimens without integrated microwires. The result reveals a clear dependence of scattering parameters with tensile force applied. Outcomes are considered as a baseline for a new detection method that can be further developed. This would allow to monitor GFRP in a non-destructive and contactless operation by embedding continuous ferromagnetic microwires within their structure.

Purpose This paper studies the feasibility of additive manufacturing (AM) processes and their potential materials to produce mold inserts for injection molding (IM) of plastic parts. This study aims to describe the technological advancements and practical implications of integrating AM with IM to produce plastic parts, reducing the gap and determining the feasibility of AM for insert production. Design/methodology/approach A literature search of different databases was done. Applying the PRISMA methodology, the 67 most relevant articles between 2013 and 2024 were selected. From these, a bibliometric analysis was performed, and the main results regarding the mechanical properties and the number of injection cycles achieved by inserts were extracted. Findings Material jetting (MJT), vat photopolymerization via ultraviolet light laser (VPP-UVL) and laser powder bed fusion using metal powders (PBF-LB/M) are the most useful AM processes reported in the literature to produce inserts for IM. Studies show that the maximum number of successful injection cycles achieved with these AM inserts has been 116, 85 and more than 500 cycles, respectively. The molded geometry, the injected material and the IM parameters influence the number of injection cycles, being the injection pressure, the mold temperature and the injection temperature the most critical parameters to consider in extending the life of the inserts. Originality/value To the best of the authors’ knowledge, this study provides the first systematic review with a comprehensive overview of this innovative approach to evaluate the emerging directions, current barriers and future potential of using AM with IM to manufacture plastic parts. As such, this study highlights the primary findings in the literature concerning AM processes and the materials commonly used to manufacture inserts for IM.