https://bura.brunel.ac.uk/handle/2438/8630 2026-08-21T08:34:11Z https://bura.brunel.ac.uk/handle/2438/33697 Title: Intelligent outlier detection under imperfect industrial data conditions Authors: Fang, Jingzhong Abstract: With the growing complexity and data intensity of modern industrial systems, intelligent data analysis (IDA) has become essential for reliable data interpretation and efficient operation. By leveraging advanced analytical and computational techniques, IDA can handle complex, largescale data sets, thereby providing valuable insights for industrial operations and significantly improving the stability and efficiency of industrial processes. Nevertheless, data quality remains a fundamental prerequisite for IDA performance, as it directly affects the accuracy and trustworthiness of derived insights. Due to complex operating conditions, high cost, and limited availability of expert annotation, high-quality data is often difficult to acquire. The imperfections in the data can affect training and evaluation, thereby reducing the overall performance in real-world applications. In practice, imperfections in industrial data generally fall into two main categories: 1) data scarcity, such as class imbalance or limited sample size; and 2) label quality issues, such as missing labels or noisy labels. In this thesis, we deal with the above-mentioned data imperfections arising from the data acquisition process and the annotation process to develop innovative approaches for robust and reliable IDA in industrial applications. It should be pointed out that all approaches developed in this thesis have been evaluated and applied to outlier detection on imperfect industrial data collected from real-world wire arc additive manufacturing (WAAM) processes. • To achieve outlier detection on unlabeled data, an improved optimization-based clustering algorithm is proposed, where the initial locations of the cluster centroids in the fuzzy C-means algorithm are optimized by an improved particle swarm optimization (PSO) algorithm. An adaptive switching randomly perturbed particle swarm optimization (ASRPPSO) algorithm is developed to enhance the convergence rate and particle’s search ability of the PSO algorithm, where a distance-based weighting strategy and switching strategy are introduced to update parameters of the optimizer. The particles can conduct a thorough search by accounting for both evolutionary states and the distances to the global and personal best positions, thereby improving the convergence rate and solution accuracy. Via the ASRPPSO-based selection of optimal clustering centroids, the proposed algorithm does not rely on centroid initialization, thereby facilitating a better cluster partition. • With the aim to guarantee the outlier detection performance on imbalanced data under the small sample problem, an optimized deep transfer learning framework is developed, where a novel deep domain adaptation strategy is designed to minimize the cross-domain discrepancies, the weighting factors are designed to handle the data imbalance problem, and the PSO algorithm is utilized for hyper-parameters tuning. By leveraging the domain knowledge and optimal hyper-parameter selection, the developed framework effectively balances performance and efficiency when dealing with outlier detection tasks on imbalanced data under the small sample problem. • To handle the noisy label problem under limited-data conditions in outlier detection, a novel Transformer-embedded learning with noisy labels framework with fuzzyclustering- assisted contrastive learning (TFCCL) is developed, where a fuzzy-clusteringassisted contrastive learning approach, a dynamic two-stage training scheme and a joint learning strategy are introduced to train the outlier detector. The TFCCL framework integrates supervised learning with contrastive learning, thereby reducing reliance on potentially noisy labels and enhancing model robustness. • For the purpose of outlier detection under the noisy label problem when sufficient data are available, a role-differentiated learning with noisy labels (RD-LNL) approach is put forward, where a leader-follower-inspired sample selection (LFSS) strategy is proposed for identifying potential clean samples for model training. A selection metric and an adaptive selection scheme are designed to adjust the number of selected samples. By combining the joint learning and adaptive sample selection, the RD-LNL framework achieves robust outlier detection in the presence of noisy labels. • To fully evaluate the application potential of the developed frameworks. All the developed frameworks are applied to outlier detection tasks on WAAM data sets collected from real-world manufacturing processes to examine their adaptability, robustness, and generalization capability. The experimental results demonstrate the effective and robust performance of the developed outlier detection frameworks on WAAM data sets across various data imperfection conditions. Description: This thesis was submitted for the award of Doctor of Philosophy and was awarded by Brunel University London 2026-01-01T00:00:00Z https://bura.brunel.ac.uk/handle/2438/33691 Title: Grasping techniques: innovations in robotics and virtual reality Authors: Zhou, Mingzhao; Aburumman, Nadine Abstract: The act of grasping is a fundamental mode of interaction when manipulating objects, in both physical and virtual environments. Robotic grasping has been studied for more than three decades, leading to the development of sophisticated frameworks. In contrast, achieving believable grasping in virtual reality (VR) requires a complex interplay of graphics, physics, and perception, where realistic haptic feedback and high-quality rendering are crucial for user immersion. This paper reviews grasping techniques in both robotics and VR, analysing VR grasping from visual and haptic perspectives and identifying its current challenges. We then compare robotic and VR grasping, showing how the relatively mature evaluation metrics of robotics can inspire potential directions for improving VR grasping and support the development of solutions that are more stable, natural, and efficient. Beyond surveying existing methods, this paper provides an outlook on upcoming challenges and opportunities in grasping research across robotics and VR. We set out both a high-level vision towards natural, and reliable grasping in real and virtual domains and priorities, including benchmark creation, robust evaluation metrics, and improvements in haptic fidelity and latency. Description: Springer Nature is providing an unedited version [Article In Press] of this manuscript to give early access to its findings. Before final publication, the manuscript will undergo further editing. Please note there may be errors present which affect the content, and all legal disclaimers apply. 2026-08-10T00:00:00Z https://bura.brunel.ac.uk/handle/2438/33687 Title: Binding Affinity Ranking at the Molecular Initiating Event (BARMIE): An open-source computational pipeline for the rapid screening of chemical interactions with steroid receptors from many species Authors: Calahorro, Fernando; Fouladi, Parsa; Pandini, Alessandro; Khushi, Matloob; Gaihre, Yogendra; Bury, Nic R Editors: Fernandez, Elias John Abstract: A challenge in ecological risk assessment is identifying the chemicals that pose the greatest threat and determining which species are most vulnerable to them. To help address this, this study has developed an in-silico open-source tool called BARMIE (Binding Affinity Ranking at the Molecular Initiating Event) to rapidly predict the chemical binding affinity of steroid receptor proteins to synthetic steroids to identify potentially vulnerable species and chemicals of concern. BARMIE was used to screen 163 teleost fish glucocorticoid receptors (GRs) for binding to the natural ligand cortisol and to 10 synthetic glucocorticoid drugs (GCs) designed to interact within the ligand-binding pocket (LBP) of GRs. BARMIE identified species from the superorder Protacanthopterygii with high-affinity GRs to synthetic GCs (e.g., vulnerable species).. BARMIE was also used to screen binding profiles of compounds in the Medicine for Malaria Venture Global Health Priority Box to rainbow trout GRs (rtGR1 and rtGR2). Of the 178 compounds, 24 and 36 bind within the LBP of rtGR1 and rtGR2, respectively. For 30 of these compounds, transactivation activity was assessed at 1µM in the presence or absence of 1µM cortisol and confirmed 2 compounds with agonistic properties (e.g., chemicals of concern) that would require further in vitro and/or in vivo studies to assess the environmental risk. BARMIE can rapidly generate predicted binding affinities for 100’s of species and chemicals as a first screen in environmental risk assessment to provide information on which substances to prioritise in downstream tests. Description: Data Availability: The code and installation instructions are available on Github https://github.com/ParsaFouladi/Barmie.; Supporting information is available online at: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0353622#sec009 . 2026-07-15T00:00:00Z https://bura.brunel.ac.uk/handle/2438/33685 Title: Spatiotemporal graph neural networks reveal conformational binding signature in protein dynamics Authors: Motta, Stefano; Santini, Gianluca; Mansoor, Samman; Nezhad, Ferdoos Hossein; Meli, Massimiliano; Pandini, Alessandro Abstract: Biomolecular function is often controlled by structural and dynamical adaptations to binding events. Although molecular dynamics (MD) simulations can capture these events at atomic resolution, separating functional signatures from stochastic noise remains challenging. Traditional methods often struggle to isolate mechanistically relevant differences across independent replicas. Here, we introduce an explainable deep learning approach that learns state-specific dynamic signatures directly from MD trajectories. By coupling a dynamic protein graph representation with group-aware contrastive learning across independent replicas, the model detects the signatures, filtering out trajectory-specific correlations. An explainable AI framework then maps the identified differences on individual residues. We demonstrate this approach by identifying “binding-ready” conformations in a T4-Lysozyme mutant, recovering the allosteric determinants of peptide recognition in the PDZ3 domain, and isolating a ligand-independent activation signature for the A2A receptor. Our GISTnet-MD method generalizes across unseen data during comparative MD analysis, translating raw trajectory differences into residue-level determinants of protein function. Description: Code availability: The code to train GISTnet-MD and perform the explainable AI pipeline is freely available on GitHub at https://github.com/MottaStefano/GISTnet-MD.; Acknowledgement: We acknowledge CINECA for the availability of high-performance computing resources as part of the agreement with the University of Milano-Bicocca and the award under the ISCRA initiative (project HP10BBT0CP). 2026-05-21T00:00:00Z