Silesian University of Technology
Recent publications
The Aerospace Integration Research Centre (AIRC) at Cranfield University offers industry and academia an open environment to explore the opportunities for efficient integration of aircraft systems. As a part of the centre, Cranfield University, Rolls-Royce, and DCA Design International jointly have developed the Future Systems Simulator (FSS) for the purpose of research and development in areas such as human factors in aviation, single-pilot operations, future cockpit design, aircraft electrification, and alternative control approaches. Utilising the state-of-the-art modularity principles in simulation technology, the FSS is built to simulate a diverse range of current and novel aircraft, enabling researchers and industry partners to conduct experiments rapidly and efficiently. Central to the requirement, a unique, user-experience-centred development and design process is implemented for the development of the FSS. This paper presents the development process of such a flight simulator with an innovative flight deck. Furthermore, the paper demonstrates the FSS’s capabilities through case studies. The cutting-edge versatility and flexibility of the FSS are demonstrated through the diverse example research case studies. In the final section, the authors provide guidance for the development of an engineering flight simulator based on lessons learned in this project.
To enhance the thermal stability of L-PBF Al-Si-Mg alloy, Fe-based metallic glass powder was mixed with Al-Si-Mg-Zr powder to fabricate a novel alloy via L-PBF. The study systematically investigated the effects of process parameters on processability and the impact of stress-relief annealing (300^{\circ }∘C–2h) on the microstructure and mechanical properties. Results showed that metallic glass addition improved surface quality. Plastic deformation was analysed using finite element and molecular dynamics methods. The as-built alloy had a cellular substructure with Al3{\rm Al}_3Al3Zr, Si, AlFeMgSi, and AlFeSi phases. Si nanoparticles and stacking faults were observed in α-Al cells. The as-built samples had yield strength (YS) of 249±4 MPa, ultimate tensile strength (UTS) of 434±6 MPa, and 6.3±0.3% elongation. Post-annealing, the alloy retained high strength and plasticity with YS of 255±3 MPa, UTS of 449±15 MPa, and 7.2±0.3% elongation, outperforming similar L-PBF Al-Si and Al-Si-Mg alloys.
The aim of this paper is to present a methodology for implementing the high-energy orbits which is still an open problem for nonlinear energy harvesters. To achieve it, this paper presents a new design of system with a flag configuration which potential function is shaped with the use of elastic elements. We have identified the lift force in FEM for wide spectrum of air velocities and used it as excitation in dimensionless mathematical model. On this basis we have conducted simulations of energy harvesting effectiveness. In the second part of the work, we focused on identifying the coexisting solutions. Due to the existence of high-energy orbits and low-energy orbits, we conducted simulations to investigate the possibility of changing the orbit. We used the Impulse Excitation Diagram here, but supplemented it with multi-colored probability distribution maps illustrating the possibility of achieving a stable orbit at given numerical values of the impulse amplitude and duration for various values of air flow velocity. The use of probability distribution maps allow to select the optimal impulse characteristics from the point of view of the energy necessary for its initiation.
The application of image analysis methods to calculate the distance from the camera to the object allows the replacement of specialized hardware devices for distance estimation. In the case of public transport, estimation of the exact position of the passenger gives also the option to determine whether the passenger is inside or outside the vehicle. In the presented work, several distance estimation methods based on typical analytical models and machine learning (ML) methods were tested using recordings from three cameras located in the minibus model. Human head detection was used instead of the entire passenger body to avoid occlusion problems. The analytical method showed worse performance than ML methods in all cases. The difference in the performance of ML models between cameras was negligible and there was no best method found. The computational time for ML models ranges from 0.35 to 100.57 ms, which should result in successful real-world applications. The developed approach can be used not only in public transport but also in all closed areas for the calculation of people or crowd density.
Objective Facial nerve palsy is a condition that carries significant consequences, especially when it occurs during adolescence. It represents a considerable clinical problem due to limited treatment options, as well as the fact that it has significant cosmetic and functional implications. In adolescents with persistent facial palsy, microsurgical procedures for dynamic facial reanimation may effectively restore facial symmetry. A commonly performed intervention is free-functioning muscle transfer. Methods The study group includes 10 pediatric patients aged 11 to 17 with diagnosed persistent facial nerve palsy, surgically treated for the restoration of mimic muscle function between 2018 and 2022. All patients included in the study underwent surgical treatment using free-functioning muscle transfer. The average observation period was 38 months (range: 21–54 mo). To objectively assess the return of function of the transferred innervated muscle flap during treatment and postoperative recovery, the House-Brackmann Facial Nerve Grading System was used. Results At an average observation period of 38 months, at least grade 3 according to the HBGS scoring system was demonstrated in all cases. Twelve months after the procedure, 6 patients (60%) were classified as grade 3, and in 4 cases (40%), the achieved symmetry was sufficient to obtain grade 2. No patient reported limitations in the adduction of the lower limb resulting from the harvest of the gracillis muscle. Conclusions Further research should focus on patient-reported outcomes, conducting studies on a large study group could lead to the development of a clinical algorithm that could facilitate the assessment and effectively assist in the treatment of facial palsy.
This study addresses computer-aided breast cancer diagnosis through a hybrid framework for breast tumor segmentation in ultrasound images. The core of the three-stage method is based on the autoencoder convolutional neural network. In the first stage, we prepare a hybrid pseudo-color image through multiple instances of fuzzy connectedness analysis with a novel distance-adapted fuzzy affinity. We produce different weight combinations to determine connectivity maps driven by particular image specifics. After the hybrid image is processed by the deep network, we adjust the segmentation outcome with the Chan-Vese active contour model. We find the idea of incorporating fuzzy connectedness into the input data preparation for deep-learning image analysis our main contribution to the study. The method is trained and validated using a combined dataset of 993 breast ultrasound images from three public collections frequently used in recent studies on breast tumor segmentation. The experiments address essential settings and hyperparameters of the method, e.g., the network architecture, input image size, and active contour setup. The tumor segmentation reaches a median Dice index of 0.86 (mean at 0.79) over the combined database. We refer our results to the most recent state-of-the-art from 2022–2023 using the same datasets, finding our model comparable in segmentation performance.
Background Transforming growth factor beta (TGFβ) is important for the morphogenesis and secretory function of the mammary gland. It is one of the main activators of the epithelial–mesenchymal transition (EMT), a process important for tissue remodeling and regeneration. It also provides cells with the plasticity to form metastases during tumor progression. Noncancerous and cancer cells respond differently to TGFβ. However, knowledge of the cellular signaling cascades triggered by TGFβ in various cell types is still limited. Methods MCF10A (noncancerous, originating from fibrotic breast tissue) and MCF7 (cancer, estrogen receptor-positive) breast epithelial cells were treated with TGFB1 directly or through conditioned media from stimulated cells. Transcriptional changes (via RNA-seq) were assessed in untreated cells and after 1–6 days of treatment. Differentially expressed genes were detected with DESeq2 and the hallmark collection was selected for gene set enrichment analysis. Results TGFB1 induces EMT in both the MCF10A and MCF7 cell lines but via slightly different mechanisms (signaling through SMAD3 is more active in MCF7 cells). Many EMT-related genes are expressed in MCF10A cells at baseline. Both cell lines respond to TGFB1 by decreasing the expression of genes involved in cell proliferation: through the repression of MYC (and the protein targets) in MCF10A cells and the activation of p63-dependent signaling in MCF7 cells ( CDKN1A and CDKN2B , which are responsible for the inhibition of cyclin-dependent kinases, are upregulated). In addition, estrogen receptor signaling is inhibited and caspase-dependent cell death is induced only in MCF7 cells. Direct incubation with TGFB1 and treatment of cells with conditioned media similarly affected transcriptional profiles. However, TGFB1-induced protein secretion is more pronounced in MCF10A cells; therefore, the signaling is propagated through conditioned media (bystander effect) more effectively in MCF10A cells than in MCF7 cells. Conclusions Estrogen receptor-positive breast cancer patients may benefit from high levels of TGFB1 expression due to the repression of estrogen receptor signaling, inhibition of proliferation, and induction of apoptosis in cancer cells. However, some TGFB1-stimulated cells may undergo EMT, which increases the risk of metastasis.
The occurrence of linear discontinuous deformations, primarily manifesting as ground steps, is becoming increasingly prevalent in mining and post-mining areas. These deformations present a significant hazard to structures, as there are no effective protective measures currently available. An important aspect of these deformations is that they can occur several decades after mining operations have ceased, making it crucial to understand their causes and conditions of formation. This paper presents a detailed case study of ground step formation that resulted in substantial damage to storage halls. Through comprehensive analyses of geological and mining conditions, combined with rigorous calculations, the study identifies the most likely factors that triggered the deformation. Notably, these factors differ from those commonly cited in the existing literature, providing a novel contribution to the research on this issue. The findings underscore the necessity for continuous monitoring and reevaluation of post-mining areas to mitigate potential risks and develop more effective protective strategies.
Nanofiber membranes receive considerable interest recently because of their distinctive structural features, facile preparation, as well as high filtering efficiency. Due to ever‐increasing air pollution, membranes made from biodegradable materials can play a crucial part in providing purified air with minimum concerns of environmental issues after the membrane's end of service life. The purpose of this systematic review is to assess the performance of biodegradable electrospun nanofibrous membrane filters toward air sub‐micron particles. To identify relevant studies, a systematic search is carried out in major scientific search engines including PubMed, Scopus, and the Web of Science. Data extraction is used to collect the necessary information on the membranes' structural properties, as well as filtration performance metrics such as efficiency, pressure drop, and quality factor. Among the electrospun membranes derived from biodegradable polymers, the polyvinyl alcohol (PVA)‐based electrospun membranes are more effective in filtration efficiency in capturing sub‐micron particles. The results highlight that these types of membranes are effective in filtration with low energy consumption, making them more apt for air purification. The use of such membranes can supply both high filtering performance and protection of the environment.
To address the challenges of predicting groundwater quality, we propose an interpretable machine learning approach. We employ advanced algorithms, including XGBoost, Random Forest, GradientBoost, and CatBoost regressors, to develop predictive models for groundwater quality. The SHapley Additive exPlanations (SHAP) method provides insights into water quality parameters’ contributions to the irrigation water quality index (IWQI). Performance metrics like RMSE, MAE, and R2R2\mathbb {R}^{2} gauge model accuracy, while feature engineering techniques, such as recursive feature elimination with cross-validation (RFECV) and permutation importance (PI), optimize the models’ performance and feature selection. This study, conducted in M’sila state, a semi-arid region heavily reliant on groundwater for agriculture, aims to support sustainable irrigation management. The results will contribute to data-driven strategies for optimizing irrigation practices, ensuring food security, and responsible water resource management.
Land degradation (LD) is the decline in a land’s functional capacity and productive potential, which includes various anthropogenic and natural drivers. This study focuses on three primary manifestations of LD including soil erosion, landslides, and rockfalls, which are the most prevalent in the Shaqlawa district. A set of 22 LD conditioning factors, encompassing curvature, lithology, aspect, river density, soil type, lineament density, river distance, elevation, road distance, length slope (LS), land use land cover (LULC), stream power index (SPI), valley depth, profile curvature, slope, solar radiation, road density, lineament distance, rainfall, topographic wetness index (TWI), plan curvature, and normalized difference vegetation index (NDVI), were integrated into the analysis. Variance inflation factors (VIF) and tolerance (TOL) values from linear regression indicate that most LD factors have acceptable levels of multicollinearity. The Information Gain Ratio (IGR) identified key variables TWI, NDVI, and lithology—as pivotal factors for predicting LD. Additionally, the study evaluated degradation factors using various machine learning (ML) algorithms, including random forest (RF), Naive Bayes, logistic regression, rotation forest, forest penalized attributes (FPA), and Fisher’s Linear discriminant analysis (FLDA). This facilitated categorizing the study area into five susceptibility categories. The FLDA model categorized the highest area under very high degradation risk at 26.72%, emphasizing the varied insights each algorithm brought to characterizing the degradation risk. Additionally, the receiver operating characteristic curves (ROC) were employed for model validation, identifying RF as the most successful model in the training dataset with an area under the curve (AUC) of 0.882, while FLDA outperformed in the testing dataset with an AUC of 0.883. The identified LD-prone areas will help land-use planners and emergency management officials apply effective mitigation strategies for similar terrains.
Several studies show that scrap tyre rubber mixed with sand is an effective and sustainable method for mitigating vibrations. The dynamic and cyclic response of this composite soil has already been investigated. However, layered sand-rubber configurations have not been considered yet. This study reports findings of resonant column tests on three types of specimen: (a) sand-only or rubber-only, (b) layered sand-rubber, and (c) sand-rubber mixtures. The analysis allowed for an evaluation of the maximum shear modulus and its degradation with strain over a wide range of confining stress and shear strain. The evolution of the damping ratio with strain was determined analogously. Effects of pre-loading and pre-straining were also considered. The results show that the behaviour of layered specimens is much more similar to that of pure rubber than to sand-rubber mixtures, with very low shear modulus values, smaller degradation of stiffness with strain and pre-loading, and higher damping. For example, at the confining stress of 100 kPa and rubber content of 0/33.3/50/67.7/100% by volume, the small strain shear moduli for sand-rubber mixtures are equal to 98.3/30.4/15.4/7.1/1.3 MPa and 98.3/3.6-4.2/2.4-2.8/2.1/1.3 MPa for sand-rubber layered specimens, depending on the arrangement of layers. A shear beam model is shown to be adequate for calculating the response of the layered specimens comprising layers of large stiffness contrast.
Cancer detection poses a significant challenge for researchers and clinical experts due to its status as the leading cause of global mortality. Early detection is crucial, but traditional cancer detection methods often rely on invasive procedures and time-consuming analyses, creating a demand for more efficient and accurate solutions. This paper addresses these challenges by utilizing automated cancer detection through AI-based techniques, specifically focusing on deep learning models. Convolutional Neural Networks (CNNs), including DenseNet121, DenseNet201, Xception, InceptionV3, MobileNetV2, NASNetLarge, NASNetMobile, InceptionResNetV2, VGG19, and ResNet152V2, are evaluated on image datasets for seven types of cancer: brain, oral, breast, kidney, Acute Lymphocytic Leukemia, lung and colon, and cervical cancer. Initially, images undergo segmentation techniques, proceeded by contour feature extraction where parameters such as perimeter, area, and epsilon are computed. The models are rigorously evaluated, with DenseNet121 achieving the highest validation accuracy as 99.94%, 0.0017 as loss, and the lowest Root Mean Square Error (RMSE) values as 0.036056 for training and 0.045826 for validation. These results revealed the capability of AI-based techniques in improving cancer detection accuracy, with DenseNet121 emerging as the most effective model in this study.
Institution pages aggregate content on ResearchGate related to an institution. The members listed on this page have self-identified as being affiliated with this institution. Publications listed on this page were identified by our algorithms as relating to this institution. This page was not created or approved by the institution. If you represent an institution and have questions about these pages or wish to report inaccurate content, you can contact us here.
6,367 members
Krzysztof Psiuk-Maksymowicz
  • Faculty of Automatic Control, Electronics and Computer Science
Henryk Palus
  • Department of Data Science and Engineering
Marek Flekiewicz
  • Department of Road Transport
Jaroslaw Figwer
  • Institute of Automatic Control
Kishore Kumar Kadimpati
  • Department of Environmental Biotechnology
Information
Address
Gliwice, Poland