Recent advances in 2D Gaussian Splatting (2DGS) have demonstrated compelling rendering efficiency and mesh extraction capabilities. However, its application to large-scale aerial photogrammetry, especially using oblique UAV imagery, remains limited due to three primary challenges: (1) suboptimal image selection in scene partitioning strategies failing to scale effectively; (2) densification pipelines that rely primarily on single-view constraints, resulting in under-reconstructions and loss of fine geometric detail; and (3) the absence of multi-view geometric consistency constraints leading to surface artifacts and inconsistencies. To address these limitations, we propose ULSR-GS, a novel method tailored for high-resolution surface reconstruction in urban-scale environments. Firstly, we propose a point-to-photo partitioning strategy that segments the scene based on the sparse SfM point cloud and assigns only the most relevant images to each sub-region, which resolves key scalability bottlenecks. Secondly, we propose a multi-view guided densification strategy that enforces adaptive geometric consistency across views, overcoming the limitations of single-view-based densifications. Lastly, we introduce consistency-aware loss functions that explicitly regulate depth and normal alignment across views, significantly enhancing surface fidelity. Extensive experiments on large-scale aerial benchmark datasets demonstrate that ULSR-GS consistently outperforms existing single-and multi-GPU Gaussian Splatting methods. Furthermore, compared to MVS pipelines, our approach achieves comparable or superior geometric quality while being substantially more time-efficient, making it a practical solution for scalable 3D modeling in digital twin and urban mapping applications. Project page: https://ulsrgs.github.io.

Dog “aggression” in the veterinary practice is commonplace. Therefore, student knowledge and education about dog behaviour and the ability to interpret “aggressive” behaviour is important from a human injury prevention and dog welfare perspective. The study aimed to compare first-year veterinary students’ perceived safest proximity to both an “aggressive” and non-reactive simulated dog, both before and after a teaching intervention about canine behaviour and a handling practical. It also examined student confidence and their ability to identify “aggressive” behaviours. Forty first year veterinary students took part in two surveys. Each survey included two videos: one of a simulated dog displaying “aggressive” behaviour, based on the ‘Canine Ladder of Aggression’; and another displaying non-reactive (passive behaviours without reaction to the participants) behaviours. Each video depicted the slow and consistent approach towards the virtual dog within a virtual indoor environment, and participants were asked to press stop if or when they would stop approaching the dog. In the “aggressive” scenario, there was a reduction in the approach-stop time from survey 1 (median = 17.8 s) to survey 2 (median = 15.2 s) in the intervention group (p = 0.018) but not in the control group (p = 0.147). Regarding confidence, there was a significant increase in the self-reported confidence rating relating to a participant’s ability to interpret canine behaviour in both the control (p = 0.011) and intervention (p = 0.003). In conclusion, these results indicate that students using approach-stop videos stayed further away from an “aggressive” virtual dog model if they had undertaken a canine behaviour educational intervention. This novel approach has the potential for further use in teaching and assessment of student knowledge and behaviour which may otherwise be difficult to demonstrate.

Coupling of nuclear codes can be performed at several scale levels and is necessary to improve their reliability and sustainability. Mainly, the coupling of nuclear codes that use nodal expansion neutronics with channel thermal hydraulics has been carried out at the fuel assembly level while, only recently, the coupling of nuclear codes that use advanced neutronics, thermal hydraulics, and thermo-mechanics has been carried out at either the fuel pin or materials level. Meanwhile, in the UK, a multiscale and multi-physics software development between NURESIM and CASL is being developed, which includes a coupling software environment that enables the coupling of nuclear codes at several scale levels. Full coupled reactor physics at either the fuel pin or materials level can be obtained by coupling the transport code LOTUS, the subchannel code CTF, and the nodal code DYN3D. In this journal article, a multi ways coupling between LOTUS and CTF at either the fuel pin or materials levels with DYN3D at the fuel assembly level is compared to a multi ways coupling between DYN3D and CTF at the fuel pin level with DYN3D at the fuel assembly level and a multi ways coupling between Open MC and CTF at the materials level with DYN3D at the fuel assembly level. These comparisons have been carried out to present the coupled reactor physics verifications at either the fuel pin or materials levels. These show that the multi ways coupling between LOTUS and CTF at either the fuel pin or materials levels with DYN3D at the fuel assembly level outperforms the multi ways coupling between DYN3D and CTF at the fuel pin level with DYN3D at the fuel assembly level due to the application in the former of full neutron transport. Also, these show that the multi ways coupling between LOTUS and CTF at the materials level with DYN3D at the fuel assembly level agrees with the multi ways coupling between Open MC and CTF at the materials level with DYN3D at the fuel assembly level due to the application in both of full neutron transport.

Building a high-precision bathymetry digital elevation model is essential for navigation planning, marine and lake resource planning, port construction, and underwater archaeological projects. However, existing bathymetry methods have yet to be effectively and comparatively analyzed. This paper comprehensively reviews state-of-the-art bathymetry methods, including data acquisition techniques, model accuracy, and interpolation algorithms for underwater terrain mapping. First, We assess the merits and drawbacks of novel data acquisition devices, such as single-beam/multi-beam echo sounders and light detection and ranging systems. After that, we analyze the accuracy of the ETOPO1, GEBCO_2022 and SRTM15 to provide valuable insights into their performance. Furthermore, we evaluate ANUDEM, Inverse Distance Weighting, Kriging and Nearest Neighbor interpolation algorithms in different underwater terrains by comparing their applicability, reliability, and accuracy in various underwater environments. Finally, we discuss the development trends and challenges in underwater bathymetry technology and offer a forward-looking perspective on the future of this essential field.

The Digital Elevation Model (DEM) super-resolution approach aims to improve the spatial resolution or detail of an existing DEM by applying techniques such as machine learning or spatial interpolation. Convolutional Neural Networks and Generative Adversarial Networks have exhibited remarkable capabilities in generating high-resolution DEMs from corresponding low-resolution inputs, significantly outperforming conventional spatial interpolation methods. Nevertheless, these current methodologies encounter substantial challenges when tasked with processing exceedingly high-resolution DEMs (256×256,512×512, or higher), specifically pertaining to the accurate restore maximum and minimum elevation values, the terrain features, and the edges of DEMs. Aiming to solve the problems of current super-resolution techniques that struggle to effectively restore topographic details and produce high-resolution DEMs that preserve coordinate information, this paper proposes an improved DEM super-resolution Transformer(DSRT) network for large-scale DEM super-resolution and account for geographic information continuity. We design a window attention module that is used to engage more elevation points in low-resolution DEMs, which can learn more terrain features from the input high-resolution DEMs. A GeoTransform module is designed to generate coordinates and projections for the DSRT network. We conduct an evaluation of the network utilizing DEMs of various types of terrains and elevation differences at resolutions of 64×64,256×256 and 512 × 512. The network demonstrated leading performance across all assessments in terms of root mean square error (RMSE) for elevation, slope, aspect, and curvature, indicating that Transformer-based deep learning networks are superior to CNNs and GANs in learning DEM features.

Dog aggression is a public health concern because dog bites often lead to physical and psychological trauma in humans. It is also a welfare concern for dogs. To prevent aggressive behaviours, it is important to understand human behaviour towards dogs and our ability to interpret signs of dog aggression. This poses ethical challenges for humans and dogs. The aim of this study was to introduce, describe and pilot test a virtual reality dog model (DAVE (Dog Assisted Virtual Environment)). The Labrador model has two different modes displaying aggressive and non-reactive non-aggressive behaviours. The aggressive behaviours displayed are based on the current understanding of canine ethology and expert feedback. The objective of the study was to test the recognition of dog behaviour and associated human approach and avoidance behaviour. Sixteen university students were recruited via an online survey to participate in a practical study, and randomly allocated to two experimental conditions, an aggressive followed by a non-reactive virtual reality model (group AN) or vice versa (group NA). Participants were instructed to 'explore the area' in each condition, followed by a survey. A Wilcoxon and Mann Whitney U test was used to compare the closest distance to the dog within and between groups respectively. Participants moved overall significantly closer to the non-reactive dog compared to the aggressive dog (p≤0.001; r = 0.8). Descriptions of the aggressive dog given by participants often used motivational or emotional terms. There was little evidence of simulator sickness and presence scores were high indicating sufficient immersion in the virtual environment. Participants appeared to perceive the dog as realistic and behaved and interacted with the dog model in a manner that might be expected during an interaction with a live dog. This study also highlights the promising results for the potential future use of virtual reality in behavioural research (i.e., human-dog interactions), education (i.e. safety around dogs) and psychological treatment (e.g. dog phobia treatment).

Many approaches are currently being considered to develop digital twins and integrated digital frameworks. These approaches concentrate on an industrial sector or a single product, with the obvious risks of a lack of interoperability at differing levels of abstraction. However, a unified approach is required to truly realize the benefits of digital twins and is achievable at their current stage of development. We present proof-of-concept case studies where a digital twin architecture has been developed using standard techniques and adopting a systems-based approach. This architecture has demonstrated technical benefits such as a step-change in processing efficiency, a reduction in traditional manual interventions, an ability to integrate risk and uncertainty, and a perceived cost benefit despite the necessary up-front investment. This approach also allows for the integration of models and simulations at differing levels of detail within an overall value chain. Several wider benefits to an organization such as improved communication and recording of implicit expert knowledge were identified. These benefits we believe will offset the upfront resource required for the development of the architecture. Challenges to adopting the technology were also identified and should be addressed in parallel with future technology development. This work is a first step in establishing a practical approach to realizing a digital twin architecture that demonstrates the flexibility and scalability that can be applied universally. We discuss the wider application of this architecture to a wide range of industry sectors by adopting this unified approach and thereby realize the major benefits a digital twin can provide.

Traditionally, the complex coupled physical phenomena in nuclear reactors has resulted in them being treated separately or, at most, simplistically coupled in between within nuclear codes. Currently, coupling software environments are allowing different types of coupling, modularizing the nuclear codes or multi-physics. Several multiscale and multi-physics software developments for LWR are incorporating these to deliver improved or full coupled reactor physics at the fuel pin level. An alternative multiscale and multi-physics nuclear software development between NURESIM and CASL is being created for the UK. The coupling between DYN3D nodal code and CTF subchannel code can be used to deliver improved coupled reactor physics at the fuel pin level. In the current journal article, the second part of the DYN3D and CTF coupling was carried out to analyse a parallel two-way coupling between these codes and, hence, the outer iterations necessary for convergence to deliver verified improved coupled reactor physics at the fuel pin level. This final verification shows that the DYN3D and CTF coupling delivers improved effective multiplication factors, fission, and feedback distributions due to the presence of crossflow and turbulent mixing.

Understanding and optimizing the relation between nuclear reactor components or physical phenomena allows us to improve the economics and safety of nuclear reactors, deliver new nuclear reactor designs, and educate nuclear staff. Such relation in the case of the reactor core is described by coupled reactor physics as heat transfer depends on energy production while energy production depends on heat transfer with almost none of the available codes providing full coupled reactor physics at the fuel pin level. A Multiscale and Multiphysics nuclear software development between NURESIM and CASL for LWRs has been proposed for the UK. Improved coupled reactor physics at the fuel pin level can be simulated through coupling nodal codes such as DYN3D as well as subchannel codes such as CTF. In this journal article, the first part of the DYN3D and CTF coupling within the Multiscale and Multiphysics software development is presented to evaluate all inner iterations within one outer iteration to provide partially verified improved coupled reactor physics at the fuel pin level. Such verification has proven that the DYN3D and CTF coupling provides improved feedback distributions over the DYN3D coupling as crossflow and turbulent mixing are present in the former.

Simulation codes allow one to reduce the high conservativism in nuclear reactor design improving the reliability and sustainability associated with nuclear power. Full-core coupled reactor physics at the rod level are not provided by most simulation codes. This has led in the UK to the development of a multiscale and multiphysics software development focused on LWRS. In terms of the thermal hydraulics, simulation codes suitable for this multiscale and multiphysics software development include the subchannel code CTF and the thermal hydraulics module FLOCAL of the nodal code DYN3D. In this journal article, CTF and FLOCAL thermal hydraulics validations and verifications within the multiscale and multiphysics software development have been performed to evaluate the accuracy and methodology available to obtain thermal hydraulics at the rod level in both simulation codes. These validations and verifications have proved that CTF is a highly accurate subchannel code for thermal hydraulics. In addition, these verifications have proved that CTF provides a wide range of crossflow and turbulent mixing methods, while FLOCAL in general provides the simplified no-crossflow method as the rest of the methods were only tested during its implementation into DYN3D.

The UK Department of Business, Energy and Industrial Strategy (BEIS) recently launched an R&D programme in Digital Reactor Design, incorporating the development of a Nuclear Virtual Engineering Capability with an integrated Modelling and Simulation programme. A key challenge of nuclear reactor design and analysis is the system complexity, which arises from a wide range of multi-physics phenomena being important across multiple length scales. This project constitutes the first step towards developing an integrated nuclear digital environment (INDE) linking together models across physical domains and incorporating real world data across all stages of the nuclear lifecycle. Simulation case studies will be developed within the INDE framework, delivering an enhanced modelling capability while ensuring the framework has immediate application. For these case studies have been specified that are relevant to design and operation phases for AGR and PWR type reactors. The AGR case considers the through-life structural performance of graphite bricks. This involves modelling of multi-scale, multi-physics phenomena in the support of reactor operations. The PWR case study is based on core multiphysics modelling, with potential relevance to operating and future PWRs, and in particular in the design of SMRs.

In collaboration with Airbus-UK, the dimensional growth of aircraft panels while being riveted with stiffeners is investigated. Small panels are used in this investigation. The stiffeners have been fastened to the panels with rivets and it has been observed that during this operation the panels expand in the longitudinal and transverse directions. It has been observed that the growth is variable and the challenge is to control the riveting process to minimize this variability. In this investigation, the assembly of the small panels and longitudinal stiffeners has been simulated using static stress and nonlinear explicit finite element models. The models have been validated against a limited set of experimental measurements; it was found that more accurate predictions of the riveting process are achieved using explicit finite element models. Yet, the static stress finite element model is more time efficient, and more practical to simulate hundreds of rivets and the stochastic nature of the process. Furthermore, through a series of numerical simulations and probabilistic analyses, the manufacturing process control parameters that influence panel growth have been identified. Alternative fastening approaches were examined and it was found that dimensional growth can be controlled by changing the design of the dies used for forming the rivets.

Purpose – The purpose of this research is to capture organisational barriers that can inhibit energy reduction in manufacturing. Energy consumption is a significant contributor to the economic and environmental components of industrial sustainability, and there is a significant body of knowledge emerging on the technical steps necessary to reduce that consumption. Achieving technical success requires organisational alignment, without which barriers to energy efficiency can be experienced. Design/methodology/approach – The research uses a theory building–theory testing cycle to propose and then verify existence of barriers to industrial energy efficiency. Literature review is used to build potential organisational barriers that can arise. The existence of barriers is then verified in industrial energy reduction projects using interview, observation and document analysis. Findings are validated by company staff. Findings – From the literature barriers that can be related to energy reduction, projects are uncovered. The generic and energy reduction-specific barriers are confirmed and two new barriers are identified. A cognitive map linking the relationships between all the barriers is proposed. Research limitations/implications – The research is built on detailed examination of a number of projects in a single company and work is needed to verify the findings in companies of different size and different industrial sector. Practical implications – The list of barriers created can support industry in preparing for and undertaking energy efficiency projects. The cognitive map proposed will help industry and academia understand why removing current prominent barriers can lead to surfacing of new barriers. Originality/value – The novelty of this research is in both the creation of a list of organisational barriers for energy efficiency as well as identifying the relationships between them. The work brings generic change management barriers to enhance the specific energy reduction barriers together into a broader collation of barriers as well as uncovering new barriers. The work proposes a cognitive map of industrial energy efficiency barriers to demonstrate their interrelationships.

Shot peening is a widely used mechanical surface processing method for fatigue strength enhancement. It is a complex process as multiple nonlinear factors are involved. In this paper, a novel method is presented to model and simulate shot peening process in 3D system. Workpieces and tools were modelled in 3D system to simulate the dynamic characteristics of the shot peening process. A vectorial routing method was introduced to define the movement of nozzles in 3D space. Shot peening impacts over the surface of the specimen was simulated with a segmented field function. The deformation of specimen was simulated instantaneously with a combined analytical/numerical method by dividing the shot peened area of the specimen into segments according to the segmented field function of coverage. Verification of simulation results by shot peening experiments showed that this approach could facilitate process design and predict operational behaviours. © Springer-Verlag London 2013.

Traditionally, manufacturing facilities and building services are analysed separately to manufacturing operations. This is despite manufacturing operations using and discarding energy with the support of facilities. Therefore improvements in energy and other resource use to work towards sustainable manufacturing have been sub-optimal. This paper presents research in which buildings, facilities and manufacturing operations are viewed as inter-related systems. The objectives are to improve overall resource efficiency and to exploit opportunities to use energy and / or waste from one process as potential inputs to other processes. The novelty here is the combined simulation of production and building energy use and waste in order to reduce overall resource consumption. The paper presents a literature review, develops the conceptual modelling approach and introduces the prototype IES Ltd THERM software. The work has been applied to industrial cases to demonstrate the ability of the prototype to support activities towards sustainable manufacturing.

The effect of shot peening on the fatigue resistance of a 7150-T651 aluminium alloy was studied using an in situ four point bending fatigue machine under an optical microscope. The tests results showed that the fatigue lives for peened specimens are longer than those for unpeened specimens, over a wide range of applied stress. They also showed that the crack growth rate of the failure crack increased steadily in unpeened specimens while in peened specimens there was an initial decrease to a minimum, occurring when the crack length was equal to the peening depth. The potential of extending the life of fatigue damaged components was investigated by shot peening specimens which had been fatigue damaged to different degrees. The results indicated that cracks shorter than the peening depth were arrested by shot peening. However, little benefits are achieved in terms of life extension by shot peening specimens with fatigue cracks longer than the peening depth. The 7150 alloy is very sensitive to the shot peening conditions used. Inappropriate peening leads to early crack formation and shorter fatigue life. A fatigue crack model incorporating shot peening effects is used to predict fatigue life of peened specimens. INTRODUCTION Of the various types of structural failures, fatigue is considered to be the most important and crucial mechanism. One method used for preventing or delaying crack initiation and propagation is to introduced a compressive stress on the material by means of shot peening. The common belief is that this treatment introduces a residual compressive stress and that it is more difficult for fatigue cracks to initiate and propagate under these conditions [1 3]. Most of the work carried out up to date on the effects of shot peening on fatigue does not consider the important aspects of initiation and propagation of cracks. However, it is well known now that fatigue crack growth in the earlier stages is affected by the microstructure and by the mechanical and work-hardening state of the material [4 7]. Shot peening, besides inducing compressive residual stresses, severely distorts the crystalline structure, which there has a marked effect on crack initiation and propagation. The major objective of this work is to determine the effect of shot peening on fatigue crack initiation and propagation in 7150-T651 aluminium alloy, which is a relatively new material widely used by the aerospace industry in the top wing skin of aircrafts. This will lead to a better understanding of the reasons behind the observed improvement in fatigue life, and to the optimisation of the shot peening process in terms of fatigue resistance. EXPERIMENTAL PROCEDURE Material The material under investigation was a 7150-T651 aluminium alloy, with an elastic modulus in the range 7175 GPa, a chemical composition of (%): Si 0.12, Fe 0.15, Cu 1.9-2.5, Mn 0.1, Mg 2-2.7, Cr -0.04, Zn 5.9-6.9, Ti 0.06, remainder Al and mechanical properties as yield strength: 450 MPa, tensile strength: 530 MPa, elongation: 2.5% [8]. Specimen preparation Aluminium 7150-T651 alloy in the form of 25 mm thick plate was used for the tests. Average sizes of the specimens were 6.5 mm thick, 4.6 mm wide and 80 mm long. Some specimens were shot peened using the process parameters given in Table 1. TABLE 1 SHOT PEENING PARAMETERS Almen intensity Shot diameter, mm Pressure, psi Shot type Time, s 14A 0.8 55 Cast steel 15 Metallography Unpeened specimens Because the 7150-T651 aluminium alloy was obtained in the form of rolled plate, the alloy had a preferred grain orientation. The average grain sizes were as follows: 28.5 μm in the longitudinal direction, 14.3 μm in the transverse direction, 9.3 μm in the short transverse direction. Shot-peened specimens The metallography revealed that the depth of the compressed layer was approximately 0.1-0.2 mm. This depth could be evaluated observing the short-transverse section of the specimen. The average grain size in the short-transverse direction was 7.8 μm, which is less than for the unpeened specimen. Loading conditions Fig. 1 shows the four point bend loading configuration used, where c = 21.8 mm and a = 22.1 mm. The four point bend facility was purposely designed for in-situ testing, using an optical fatigue damage viewing facility. It provides the possibility of continuously monitoring crack propagation during a fatigue test. It is a simple matter to stop the test, scan the specimen gauge surface to find points of interest and then, when a crack has formed, continuously monitor it up to failure or until the crack has reached the required length. A frequency of 15 Hz and a stress ratio of 0.1 were used in the tests. The specimens loading conditions are listed in Table 2.

Traditionally, manufacturing facilities and building services are considered to be supplementary to manufacturing operations. Manufacturing operations use and discard energy with the support of facilities. Additionally, these disciplines have been pursued independently to date; the design and improvement of buildings and manufacturing systems typically carried out in isolation. Therefore improvements in energy and other resource use to work towards sustainable manufacturing have been sub-optimal. This paper presents research in which buildings, facilities and manufacturing operations are viewed as interrelated systems. The work documents practices, tactics and specifies modeling tools, that enable integrated analysis of manufacturing site energy and resource use. The objectives are to improve overall energy efficiency and to exploit opportunities to use energy or other waste from one process as potential inputs to other processes. The novelty here is the combined simulation of production and building energy use and waste in order to reduce overall resource consumption. The paper presents a literature review, develops the conceptual modeling approach and introduces the prototype with industrial cases to support activities towards sustainable manufacturing. Introduction The Bruntland report (WCED 1987) defines sustainable development as meeting the needs of the present generation without compromising the ability of future generations to meet their own needs. Focusing on sustainable manufacturing there is the need to recognise the triple bottom line of social justice (people), environmental quality (planet) and economic prosperity (profit) (Elkington, 1997). There is significant work underway in both academia and industry to develop tools and techniques for sustainable manufacturing and apply them for tangible benefit. As a result of rising prices and concerns over energy security and climate change, energy is a major focus. Documented cases and achievements presented on corporate websites show that significant benefits can be obtained. However, it is not until a sustainability mindset is adopted that the opportunities can be identified in the first place.

Growing environmental concerns caused by natural resource depletion and pollution need to be addressed. One approach to these problems is Sustainable Development, a key concept for our society to meet present as well as future needs worldwide. Manufacturing clearly has a major role to play in the move towards a more sustainable society. However it appears that basic principles of environmental sustainability are not systematically applied, with practice tending to focus on local improvements. The aim of the work presented in this paper is to adopt a more holistic view of the factory unit to enable opportunities for wider improvement. This research analyses environmental principles and industrial practice to develop a conceptual manufacturing ecosystem model as a foundation to improve environmental performance. The model developed focuses on material, energy and waste flows to better understand the interactions between manufacturing operations, supporting facilities and surrounding buildings. The research was conducted in three steps: (1) existing concepts and models for industrial sustainability were reviewed and environmental practices in manufacturing were collected and analysed; (2) gaps in knowledge and practice were identified; (3) the outcome is a manufacturing ecosystem model based on industrial ecology (IE). This conceptual model has novelty in detailing IE application at factory level and integrating all resource flows. The work is a base on which to build quantitative modelling tools to seek integrated solutions for lower resource input, higher resource productivity, fewer wastes and emissions, and lower operating cost within the boundary of a factory unit.