The radome of weather radars can be covered with a layer of water, degrading the quality of the radar products. Considering a simplified setup with a planar replica of the Swiss weather radars’ radome, we measure and model analytically its scattering parameters, with and without water. The measured reflectance of the dry radome replica agrees well with the one modeled according to the manufacturer specifications. Water forms droplets on the hydrophobic surface, but water films thicker than 1 mm can be created. Meteorologically more realistic thinner water films are expected on old radomes that have become hydrophilic with aging. Using hygroscopic silk and cotton tissues, we empirically imitate water films as thin as less than 0.1 mm. The measurements align with the simple analytical model of uniform plane wave incidence on the radome and water film but could be further improved by taking refraction and bending of the radome replica into account. Simulations with the General Reflector Antenna Software Package (GRASP) from TICRA complement the study for a representative setup with a spherical radome.

We investigated whether executive functions (EFs) are engaged in bilingual language control in Finnish speakers with different degrees of Swedish language experience and proficiency, including early bilinguals, late high-proficiency bilinguals and low-proficiency learners of Swedish. In an online experiment, language switching was measured with a cued naming (CN) paradigm, and a Simon task was used to assess EF performance. Following the skill-learning (task specificity) hypothesis, we expected that language switching may be automatized and no longer rely on EFs in bilinguals with high language proficiency, but not for those with lower proficiency. Thus, we expected significant associations between the tasks in the lower proficiency participants only. Our results showed no CN switching–EF associations in the more experienced L2 speakers, but a significant association in lower-proficiency participants. This suggests that language switching engages EFs only in participants with lower proficiency in whom these processes are not yet automatized.

Child and youth mental health is an international public health and research priority. We are an interdisciplinary and cross-sectoral network of UK-based early career researchers (ECRs) with an interest in child and youth mental health research. In this paper, we reflect on ongoing challenges and areas for growth, offering recommendations for key stakeholders in our field, including researchers, institutions and funders. We present a vision from an ECR perspective of what future child and youth mental health research could look like and we explore how the research infrastructure can support ECRs and the wider research field in making this vision a reality. We focus specifically on: (a) embracing complexity; (b) centring diverse voices; and (c) facilitating sustainable research environments and funding systems. We present recommendations for all key partners to consider alongside their local contexts and communities to actively and collaboratively drive progress and transformative change.

The article examines the relationship between perceived distributive justice and trust in the welfare system within complex and self-contradictory policy setting. Based on thirty-three in-depth interviews with social assistance users in Poland and Czechia, we find that policy assemblages in those countries are experienced as confusing ‘institutional enigmas’. We identify four patterns linking perceptions of welfare system’s distributive justice and trust in this context: perceived rationality of the system combined with trust; perceived lack of system’s empathy combined with distrust; concerns about ‘undeserving claimants’ overusing the system linked to distrust in welfare system; and unexpected (non)receiving of benefits causing surprise and shaping (dis)trust. We argue that in contradictory institutional embedding, achieving users’ trust is challenging due to complex distributive justice principles they adhere to and numerous instances of those principles being violated. Trust can still be fostered when users are well informed or experience receiving meaningful support.

Neurotransmitter release via synaptic vesicle fusion with the plasma membrane is driven by SNARE proteins (Synaptobrevin, Syntaxin, and SNAP-25) and accessory proteins (Synaptotagmin, Complexin, Munc13, and Munc18). While extensively studied experimentally, the precise mechanisms and dynamics remain elusive due to spatiotemporal limitations. Molecular dynamics (MD) simulations—both all-atom (AA) and coarse-grained (CG)—bridge these gaps by capturing fusion dynamics beyond experimental resolution. This review explores the use of these simulations in understanding SNARE-mediated membrane fusion and its regulation by Synaptotagmin and Complexin. We first examine two competing hypotheses regarding the driving force of fusion: (1) SNARE zippering transducing energy through rigid juxtamembrane domains (JMDs) and (2) SNAREs generating entropic forces via flexible JMDs. Despite different origins of forces, the conserved fusion pathway – from membrane adhesion to stalk and fusion pore (FP) formation – emerges across models. We also highlight the critical role of SNARE transmembrane domains (TMDs) and their regulation by post-translational modifications like palmitoylation in fast fusion. Further, we review Ca²⁺-dependent interactions of Synaptotagmin’s C2 domains with lipids and SNAREs at the primary and tripartite interfaces, and how these interactions regulate fusion timing. Complexin’s role in clamping spontaneous fusion while facilitating evoked release via its central and accessory helices is also discussed. We present a case study leveraging AA and CG simulations to investigate ion selectivity in FPs, balancing timescale and accuracy. We conclude with the limitations in current simulations and using AI tools to construct complete fusion machinery and explore isoform-specific functions in fusion machinery.

This paper examines in what way providers of specialized Large Language Models (LLM) pre-trained and/or fine-tuned on medical data, conduct risk management, define, estimate, mitigate and monitor safety risks under the EU Medical Device Regulation (MDR). Using the example of an Artificial Intelligence (AI)-based medical device for lung cancer detection, we review the current risk management process in the MDR entailing a “forward-walking” approach for providers articulating the medical device’s clear intended use, and moving on sequentially along the definition, mitigation, and monitoring of risks. We note that the forward-walking approach clashes with the MDR requirement for articulating an intended use, as well as circumvents providers reasoning around the risks of specialised LLMs. The forward-walking approach inadvertently introduces different intended users, new hazards for risk control and use cases, producing unclear and incomplete risk management for the safety of LLMs. Our contribution is that the MDR risk management framework requires a backward-walking logic. This concept, similar to the notion of “backward-reasoning” in computer science, entails sub-goals for providers to examine a system’s intended user(s), risks of new hazards and different use cases and then reason around the task-specific options, inherent risks at scale and trade-offs for risk management.

The present paper provides a small–scale exploratory analysis of L2 English pronunciation and accent aims among secondary school students in Germany – with a focus on the bath and lot vowels, rhoticity, and T–flapping. The eight learners investigated in the current study show blended use of Standard Southern British English (StSBrE) and Standard American (StAmE) phonological variants with relatively high degrees of variation between learners. StSBrE–oriented productions were dominant overall. Agreement of accent aim and L2 pronunciation was largely feature–dependent and limited overall but varied between learners.

Participatory Design – an iterative, flexible design process that closely involves stakeholders, often end users – is growing in use across design disciplines. As more practitioners use Participatory Design (PD), it has become less rigidly defined, with stakeholders engaged to varying degrees through disjointed techniques. This ambiguity can be counterproductive when discussing PD processes. We performed a systematic literature review that builds shared, foundational knowledge of PD processes and techniques while also summarizing the state of PD research in the field, as a first step in supporting richer understandings of how best to equitably engage with stakeholders. We found that a majority of PD literature examined specific case studies of PD, with the design of intangible systems representing the most common design context. Stakeholders most often participated throughout multiple stages of a design process, recruited in a variety of ways, and engaged in several of the 14 specific participatory techniques identified. Our findings also identify leverage points for creators of PD processes and how the leverage points impact design equity, including: (1) emergent versus predetermined processes; (2) direct versus indirect participation; (3) early versus late participation; (4) one time versus iterative participation; and (5) singular versus multiple PD techniques.

What can we learn about organizational ethics from studying cemeteries as organizational/organized manifestations of our mutual, embodied vulnerability? How does, and how should, the ethico-political imperative of death and the deceased materialize in the cemeterial space? With reference to a comparative analysis of two island cemeteries, Venice’s San Michele and New York’s Hart Island, this paper makes three contributions to the emerging literature on organizational ethics of life and death. First, it makes an empirical contribution based on an organizational study of two “resting places” that highlights the importance of understanding organizational life and death with reference to ethics. Second, it makes a theoretical contribution to scholarship on the organization of death and on grieving as embedded in a politics and ethics of recognition. Third, the paper shows how our desire to be recognized as valid, viable subjects comes to be organized, and situated, in ways that perpetuate precarity and vulnerability, a point that is illustrated with reference to cemeteries as ethically significant organizational settings.

Heating, Ventilation, and Air Conditioning (HVAC) systems are major energy consumers in buildings, challenging the balance between efficiency and occupant comfort. While prior research explored generative AI for HVAC control in simulations, real-world validation remained scarce. This study addresses this gap by designing, deploying, and evaluating “Office-in-the-Loop,” a novel cyber-physical system leveraging generative AI within an operational office setting. Capitalizing on multimodal foundation models and Agentic AI, our system integrates real-time environmental sensor data (temperature, occupancy, etc.), occupants’ subjective thermal comfort feedback, and historical context as input prompts for the generative AI to dynamically predict optimal HVAC temperature setpoints. Extensive real-world experiments demonstrate significant energy savings (up to 47.92%) while simultaneously improving comfort (up to 26.36%) compared to baseline operation. Regression analysis confirmed the robustness of our approach against confounding variables like outdoor conditions and occupancy levels. Furthermore, we introduce Data-Driven Reasoning using Agentic AI, finding that prompting the AI for data-grounded rationales significantly enhances prediction stability and enables the inference of system dynamics and cost functions, bypassing the need for traditional reinforcement learning paradigms. This work bridges simulation and reality, showcasing generative AI’s potential for efficient, comfortable building environments and indicating future scalability to large systems like data centers.