Digitala Vetenskapliga Arkivet - Academic Archive On-line
Not a member yet
734120 research outputs found
Sort by
Cloud-based IoT and Collaborative Learning for Cyber-Physical System of Systems
No full textThe growth in cyber-physical systems (CPS), the industrial internet of things (IIoT) and integrations of machine learning (ML) models have enabled Industry 4.0 automation and intelligence in industrial System of Systems (SoS). However, scalable automation frameworks, dynamic interoperability for smooth communication among heterogeneous systems, and the integration of ML models particularly for online collaborative intelligence in distributed architectures, are open challenges that must be addressed. This thesis structures its research as a progressive investigation, with each identified challenge leading to the next. First, local cloud-based automation is explored using the Eclipse Arrowhead framework to propose digitalization frameworks for industrial use cases, such as predictive maintenance in wind energy systems and smart manufacturing. The primary objective is to bridge the digital divide in SoSs environments, ensuring seamless intercommunication between IoT-connected devices. Significant engineering effort is dedicated to developing dynamically scalable automation solutions that integrate heterogeneous CPS, providing real-time adaptability and efficiency. This leads to the second research challenge addressed in this thesis. To investigate the challenge of semantic interoperability among heterogeneous IIoT devices, this research explores ontology alignment techniques through Natural Language Processing (NLP) and deep learning models. Extension of an existing language model (BERT_Intereaction) is proposed for ontology graphs to facilitate seamless communication between heterogeneous IIoT devices. It is designed using a language encoder to develop knowledge of the text to understand the labels or entities and a structural encoder to understand the context or semantics behind the text. This proposed model consistently outperforms cross-lingual tasks over the state-of-theart techniques with an error reduction of 2.1% on benchmark datasets DBP15KZH−EN, DBP15KJA−EN, and DBP15KFR−EN. The third challenge involves scaling collaborative intelligence across distributed IioT systems. To address this, a local cloud-based collaborative learning (CCL) model is designed for the service-oriented architecture (SOA) and a decentralized ML model to digitalize IIoTs while enabling ML-based optimizations for IIoT tasks across cloud and edge nodes. The CCL model integrates machine learning-as-a-service (MLaaS) into the distributed cloud architectures. CCL offers scalable, privacy-driven, self-contained local clouds for every CPS in the system of systems model. The local clouds enable distributed ML deployment across the IIoT SoS, where devices collaborate to share their knowledge representations. The model uses unsupervised dictionary learning, allowing IIoT nodes to share compressed, optimized learning representations. Furthermore, this thesis also highlights a culminating issue for designing decentralized ML-enabled IIoT solutions that is the information overload and redundancy at the edge and cloud. To mitigate this challenge, the CCL+ model is proposed, integrating coherence-based dictionary refinement with Bayesian optimization. The model is tested on the condition monitoring task using data from an automated farm of six wind turbines using the CCL model. Implementing redundancy-aware strategies in the CCL+ optimized bandwidth usage and reduced communication overhead, especially for the resource-constrained IIoT devices. In the simulation experiments, over one year, the propagated learned dictionary size at a single wind turbine exceeded 1 petabyte. In contrast, in comparison, using the CCL+ model, for the same duration, the learned dictionary remained at 18 MB, significantly enhancing communication and computational efficiency without losing essential information. Various potential future research directions accompany the findings presented in this thesis. For instance, to strengthen the ablation study on the semantic interoperability among heterogeneous SoS challenge, a generalized IIoT ontology that is designed for any IoT device (beyond sensors), such as the smart applications reference ontology (SAREF) can be tested for ontology alignment. This work provides a step towards enabling translation between heterogeneous IoT sensor devices. The proposed BERT Intereaction model can be further extended to a translation module using the generalized ontology graphs. Then investigations can be conducted to test if the model can interpret the messages transmitted across ontologically different devices in two scenarios: a) where the ontology graphs of both devices are a subset of the generalized ontology graph, and b) they are overlapping graphs and may contain different nodes and relations but they are semantically the same. Furthermore, exploring the utility of CCL+ model for extended large-scale SoS with multiple parallel tasks to test the collaborative learning concept across the heterogeneous cyber-physical system of systems (CPSoS)
Quantitative approaches for spatial metabolomics with isomer differentiation using surface sampling capillary electrophoresis mass spectrometry
No full textThe importance of metabolites and their isomeric structures in biological function and dysfunction is increasingly recognized. However, achieving quantitative mapping of metabolites within tissue regions, particularly with isomeric specificity, remains an analytical challenge. This work presents the development of a quantitative surface sampling capillary electrophoresis method for spatial metabolomics with isomeric resolution. Five quantitation strategies were evaluated, with the optimal approach identified as sequential injection of metabolites directly from tissue alongside standards. This methodology was applied to a rat brain tissue section in a proof-of-principle study, enabling quantitative spatial analysis of metabolites, neurotransmitters, and isomeric species. Among the findings, the aromatic amino acids tyrosine, phenylalanine, and tryptophan exhibited the most dynamic distributions across four brain regions, while leucine and isoleucine demonstrated distinct spatial profiles, with leucine consistently being the more abundant isomer. This method offers a promising tool for advancing the understanding of spatially resolved biochemical processes underlying biological function and dysfunction.Title in the list of papers of Anastasia Golubova's thesis: Quantitative spatial metabolomics with isomer differentiation using Surface Sampling Capillary Electrophoresis Mass Spectrometry</p
Differences in Drivers’ Attitudes and Behavior Towards Ambulances and Police Vehicles
No full textEmergency driving entails challenges, one being the interaction with other road users. Drivers’ attitudes towards Emergency vehicles (EVs) may affect their willingness to move over. Previous research has suggested that attitudes towards the different emergency response organizations may differ, but no previous research has compared the driver behavior aspects associated with the two organizations. The purpose of the present study was to examine drivers’ attitudes and driving behavior around approaching ambulances and police vehicles. A total of 110 drivers participated in a driving simulator study where they were approached by emergency vehicles three times. After the drive, they were asked about their typical driving behavior and attitudes towards ambulance and police. No difference in driving behavior when interacting with either an ambulance or police vehicle was seen in the simulator data. There was a difference in expressed attitude towards the organizations. Drivers reported that they felt a higher level of stress, and were more aware of their driving when approached by a police vehicle compared to when approached by an ambulance. The participants stated that it is important to move over regardless of EV type.
Near triple arrays
No full textWe introduce near triple arrays as binary row-column designs with at most two consecutive values for the replication numbers of symbols, for the intersection sizes of pairs of rows, pairs of columns and pairs of a row and a column. Near triple arrays form a common generalization of such well-studied classes of designs as triple arrays, (near) Youden rectangles and Latin squares. We enumerate near triple arrays for a range of small parameter sets and show that they exist in the vast majority of the cases considered. As a byproduct, we obtain the first complete enumerations of 6×10 triple arrays on 15 symbols, 7×8 triple arrays on 14 symbols and 5×16 triple arrays on 20 symbols. Next, we give several constructions for families of near triple arrays, and e.g. show that near triple arrays with 3 rows and at least 6 columns exist for any number of symbols. Finally, we investigate a duality between row and column intersection sizes of a row-column design, and covering numbers for pairs of symbols by rows and columns. These duality results are used to obtain necessary conditions for the existence of near triple arrays. This duality also provides a new unified approach to earlier results on triple arrays and balanced grids
Gender biases in surgical residency and their association with postoperative outcomes : a qualitative study of residents in general and orthopedic surgery
No full textObjectives: Recent studies have shown gender-based disparities in surgical outcomes, with women physicians achieving better postoperative results. However, the mechanisms underlying these differences remain poorly understood. This complex, underexplored phenomenon lacks perspectives from key stakeholders such as patients, staff, and surgeons. Surgical residents inhabit a unique position, balancing the roles of learners and healthcare providers and offering valuable insights into these disparities. Our objective was to explore resident surgeons' perceptions of the mechanisms underlying the differences observed in postoperative outcomes through interviews. Methods: We conducted a qualitative study using semi-structured interviews with men and women residents, analyzed through qualitative content analysis. Participants included 17 surgical residents from 10 ACGME-accredited orthopedic and general surgery programs across four U.S. regions. Results: Key findings were: (1) Women surgical residents experienced gender biases during training; (2) Men and women surgeons differed in their approaches to patient care and surgical practice. Residents of both genders noted that women surgeons encounter biases stemming from prevailing norms, beliefs, and expectations related to gender. These biases manifest as double standards and pressure to conform to men-dominated environments. The absence of comparable pressure on men surgeons to adapt to interactions with women colleagues or patients may lead to challenges in meeting the expectations of women patients. Participants of both genders reported that women surgeons tend to adopt a more holistic approach to patient care. Conclusions: Our findings suggest that the unique challenges faced by women surgeons may contribute to the development of refined interpersonal and surgical skills, which could be associated with improved postoperative outcomes. Raising awareness of gender-based differences in surgical practice is essential to addressing the disparities observed in postoperative results. Practical Implications: Residency programs may benefit from incorporating gender-aware training to address implicit biases, foster equitable learning environments, and optimize skill development for all trainees
Field Performance of an EcoVault Facility for Stormwater Quality Treatment
No full textAs urbanization accelerates, stormwater management in cities has shifted from focusing strictly on water quantity to addressing water quality. Traditionally implemented systems, such as stormwater ponds, while offering effective solutions, often require large land areas to implement, making them impractical for dense urban environments. Underground stormwater systems, like EcoVault, offer a more compact solution; however, they lack scientific studies under real-world conditions to prove their effectiveness in treating pollutants. This study evaluates the treatment performance of two parallel EcoVault systems with the same design, consisting of a sedimentation step and a filtration step. These facilities were retrofitted into two different stormwater sewer networks draining two urban catchments. The systems were assessed for their ability to treat total suspended solids, metals, nutrients, and organic pollutants from urban runoff. Over 15 rain events, the average total suspended solids (TSS) removal rate was 40% for EcoVault A and 46% for EcoVault B. The removal rates for metals varied, with EcoVault B showing better performance for average metal treatment (53% for Cu and 58% for Zn). However, neither EcoVault system removed dissolved metals, often with an increase of dissolved metal concentration in the effluent. The filtration step did not contribute to pollutant treatment, likely due to clogging and high hydraulic loading rates. The study highlighted the potential of underground stormwater treatment in areas with limited space availability, while identifying challenges such as treatment of dissolved pollutants.Validerad;2025;Nivå 1;2025-11-11 (u2);Full text: CC BY license;This article has previously appeared as a manuscript in a thesis.</p
Diffusion of Affibody molecules in extracellular matrix mimetic hydrogels and the effect of albumin binding
No full textAffibody molecules are protein ligands, that due to their small size (6-19 kDa) and high target affinity exhibit favourable properties for tumour uptake valuable in diagnostic imaging and therapeutic applications. Fusion to a high affinity albumin binding domain (ABD) has been shown to improve circulatory half-life and biodistribution. However, the effect of molecular design is not obvious to predict and in vitro methods to evaluate their transport properties in physiologically relevant environment are needed. In this work we investigated the diffusivities (D) of Affibody molecules, with systematically varied molecular design, in solution and within extracellular matrix mimetic hydrogels composed of either agarose or collagen and hyaluronic acid (COL-HA) using fluorescence recovery after photobleaching. Furthermore, the effect of presence of human serum albumin (HSA) was evaluated. The correlation between D of the tested Affibody molecules in solution and their molecular weight (Mw) was weak, indicating that propensity to form reversible oligomers and the size of the oligomers are more important for their diffusion properties than Mw of the monomer. Positively charged Affibody molecules were enriched in polymer-rich domains of the COL-HA gel accompanied by a decrease in D as a result of electrostatic interactions. Binding to HSA by Affibody molecules containing an ABD was evident as a decrease of D when HSA was present. In COL-HA gels HSA-binding reduced the effect of electrostatic interactions effectively facilitating the transport of those compounds. In conclusion, molecular design especially inclusion of an ABD affected the transport properties of the tested Affibody molecules
Associations between healthcare workers' substance use and quality of care : findings from a one-year Swedish follow-up study
No full textBackground: Problem drinking and illicit drug use among healthcare workers (i.e., physicians and nurses) may impair their attention and cognitive functioning, thereby increasing the risk of medical errors and diminishing the quality of patient care. Objective: To investigate the association between healthcare workers' problem drinking and illicit drug use with subsequent self-rated quality of care provided. Design: A two-wave longitudinal observational study. Methods: Panel data were drawn from the Longitudinal Occupational Health Survey in Healthcare Sweden (LOHHCS), collected in 2022 (baseline) and 2023 (follow-up), encompassing a sample of 3280 healthcare workers. Questionnaires included problem drinking, illicit drug use (cannabis and stimulants), and self-rated quality of care they provide to patients. Logistic regression models analysed relationships between the study variables. Results: At baseline, the prevalence of problem drinking was 3.8 %, and illicit drug use was 1.3 %. Both problem drinking (OR = 1.93, 95 % CI = 1.28–3.02) and illicit drug use (OR = 2.07, 95 % CI = 1.00–4.29) were significantly associated with lower self-rated quality of care provided at follow-up, after adjustment for confounding variables. Conclusions: This novel longitudinal study shows that healthcare workers reporting substance use at baseline were about twice as likely to report providing poor quality of care one year later. These findings are of clinical relevance and highlight the need for targeted preventive measures and interventions to safeguard the health and well-being of healthcare workers while maintaining quality standards in patient care
Angular-dependent interatomic potential for large-scale simulation of bcc and hcp multi-component refractory alloys
No full textThis work is devoted to the development and comprehensive validation of a new interatomic potential for bcc and hcp refractory alloys based on the W-Mo-Nb-Ta-Zr-Ti system. The presented model allows the simulation of various structural transformations, as well as the behavior of crystal defects in several of the phases observed in this system. The classical form of the potential enables simulations of atomic systems comprising up to 108 atoms for durations longer than a million time steps using a routine computational setting. The wide applicability of the developed model is demonstrated by the example of studying phase transformations in Ti Nb alloys and the properties of defects in Laves phases
Estimating the effect of cold forming and post processing on fatigue of high strength steels under component-related loading conditions
No full textStamping and shot peening of chassis components such as wheels and cross beams introduce residual stresses that affect the fatigue life. The tensile residual stresses from stamping are often found in the most critical areas for fatigue. These areas are commonly subject to bending stress states while smooth material fatigue data for steel sheets more commonly is obtained by uniaxial testing. Both the sensitivity to residual stresses and change in load condition are material dependent posing a challenge in fatigue life estimation of formed and shot-peened specimens. This paper aims to provide a convenient tool for high cycle fatigue life estimations in these conditions solely dependent on uniaxial tensile properties without parameter fitting. The underlying causes for the material dependency is discussed and reflected in the methodology. Uniaxial fatigue testing and fatigue testing of formed specimens with and without subsequent shot peening are performed for validation.Validerad;2025;Nivå 2;2025-11-24 (u4);Fulltext license: CC BY</p