University of Essex

Multi-robot system for manufacturing is an Industry Internet of Things (IIoT) paradigm with significant operational cost savings and productivity improvement, where Unmanned Aerial Vehicles (UAVs) are employed to control and implement collaborative productions without human intervention. This mission-critical system relies on 3-Dimension (3-D) scene recognition to improve operation accuracy in the production line and autonomous piloting. However, implementing 3-D point cloud learning, such as Pointnet, is challenging due to limited sensing and computing resources equipped with UAVs. Therefore, we propose a Digital Twin (DT) empowered Knowledge Distillation (KD) method to generate several lightweight learning models and select the optimal model to deploy on UAVs. With a digital replica of the UAVs preserved at the edge server, the DT system controls the model sharing network topology and learning model structure to improve recognition accuracy further. Moreover, we employ network calculus to formulate and solve the model sharing configuration problem toward minimal resource consumption, as well as convergence. Simulation experiments are conducted over a popular point cloud dataset to evaluate the proposed scheme. Experiment results show that the proposed model sharing scheme outperforms the individual model in terms of computing resource consumption and recognition accuracy.

Molecular communication (MC) is a significant technology in the field of nanobiology, which uses molecules as message carriers to transmit information. Diffusion channel model is the most commonly channel model base on Brownian motion in molecular communication. In single-input single-output (SISO) molecular communication model, inter-symbol interference (ISI) exists due to the long tail effect. In this study, inspired by the D-MoSK modulation scheme, where different types molecules used for encoding, A new simple modulation is proposed, which can not only reduce the ISI interference effectively, but also improve the transmission rate to a certain extent. Numerical results show that compared with the current modulation scheme, the proposed scheme makes the system achieve better BER performance and the transmission rate is also improved.

In a neoliberal academia dominated by masculine ideals of measurement and performance, it is becoming more important than ever to develop alternative ways of researching and writing. This powerful new book gives voice to non-conforming narratives, suggesting innova-tive, messy and nuanced ways of organizing the reading and writing of scholarship in manage-ment and organization studies. In doing so it spotlights how different methods and ap-proaches can represent voices of inequality and reveal previously silenced topics. Informed by feminist and critical perspectives, this will be an invaluable resource for current and future scholars in management and organ-ization studies and other social sciences.

The paper proposes a deep learning model based on Chebyshev Network Gated Recurrent Units, which is called Spectral Graph Convolution Recurrent Neural Network, for multichannel electroencephalogram emotion recognition. First, in this paper, an adjacency matrix capturing the local relationships among electroencephalogram channels is established based on the cosine similarity of the spatial locations of electroencephalogram electrodes. The training efficiency is improved by utilizing the computational speed of the cosine distance. This advantage enables our method to have the potential for real-time emotion recognition, allowing for fast and accurate emotion classification in real-time application scenarios. Secondly, the spatial and temporal dependence of the Spectral Graph Convolution Recurrent Neural Network for capturing electroencephalogram sequences is established based on the characteristics of the Chebyshev network and Gated Recurrent Units to extract the spatial and temporal features of electroencephalogram sequences. The proposed model was tested on the publicly accessible dataset DEAP. Its average recognition accuracy is 88%, 89.5%, and 89.7% for valence, arousal, and dominance, respectively. The experiment results demonstrated that the Spectral Graph Convolution Recurrent Neural Network method performed better than current models for electroencephalogram emotion identification. This model has broad applicability and holds potential for use in real-time emotion recognition scenarios.

Beta‐decay half‐lives for the free neutron, for ⁶ He and ⁸ He, and for ⁸ Li are calculated ab initio from geometrical thermodynamics arguments, independently of any quantum mechanics. Half‐lives for the decay of ⁸ Be to two alphas and for the disintegration of the tetraneutron are also calculated. The calculated values are close to those experimentally observed.

The article reflects upon the observational practices and methods developed by the early exponents of ethology committed to naturalistic field study and explores how their approaches and techniques influenced a wider field of popular natural-history filmmaking and photography. In doing so, my focus is upon three aspects of ethological field studies: the socio-technical devices used by ethologists to bring birds closer to them, the distinctive observational and representational practices which they forged, and the analogies they used to codify behaviour. This assemblage of elements included hides or screens from which to watch wild birds without disturbing them, optics to extend human vision, pens and paper to sketch and fix patterns of behaviour, watches to record timings, photography to capture action and freeze movement, and illustration and photographs to visualize behaviour. Carried through natural-history networks, the practices, methods and theories of ethologists like Huxley and Tinbergen influenced popular natural-history filmmaking and photography more broadly from the 1940s, driving a behavioural turn in these cultural practices. This popularization of the ‘ethological eye’ was further facilitated by the convergence of socio-technical devices, forms of observation and dramatization in the work of the early exponents of naturalistic field studies of birds and the popular filmmakers.

Background and Objective A network analysis can be used to quantitatively assess and graphically describe multiple interactions. This study applied network analyses to determine the interaction between physical and pain‐related factors and fear of movement in people with whiplash‐associated disorders (WAD) during periods of acute and chronic pain. Methods Physical measurements, including pressure pain‐thresholds (PPT) over neural structures, cervical range of motion, neck flexor and extensor endurance and the cranio‐cervical flexion test (CCFT), in addition to subjective reports including the Tampa Scale of Kinesiophobia (TSK‐11), Neck Disability Index (NDI) and neck pain and headache intensity, were assessed at baseline in 47 participants with acute WAD. TSK‐11, NDI and pain intensity were assessed for the same participants 6 months later ( n = 45). Two network analyses were conducted to estimate the associations between features at baseline and at 6 months and their centrality indices. Results Both network analyses revealed that the greatest weight indices were found for NDI and CCFT at baseline and for neck pain and headache intensity and NDI and TSK‐11 at both time points. Associations were also found betweeen cervical muscle endurance and neck pain intensity in the acute phase. Cervical muscle endurance assesssed during the acute phase was also associated with NDI after 6 months ‐ whereas PPT measured at baseline was associsated with headache intensity after 6 months. Conclusion The strongest associations were found for headache and neck pain intensity and neck disability and fear of movement, both during acute pain and when mesured 6 months later. The extent of neck endurance and measures of PPT at baseline may be associated with neck disability and headache, respectively, 6 months after a whiplash injury. Significance Through two network analyses, we evaluated the interaction between pain‐related factors, fear of movement, neck disability and physical factors in people who had experienced a whiplash injury. We demonstrated that physical factors may be involved in the maintenance and development of chronic pain after a whiplash injury. Nevertheless, the strongest associations were found for headache and neck pain intensity and neck disability and fear of movement, both during acute and chronic phases.

Objective To investigate the effects of being born late preterm (LPT, 34–36 weeks’ gestation) or early term (37–38 weeks) on children’s educational achievement between ages 5 and 11 years. Design A series of observational studies of longitudinal linked health and education data. Setting The Born-in-Bradford (BiB) birth cohort study, which recruited mothers during pregnancy between 2007 and 2011. Participants The participants are children born between 2007 and 2011. Children with missing data, looked-after-children, multiple births and births post-term were excluded. The sample size varies by age according to amount of missing data, from 7860 children at age 5 years to 2386 at age 11 years (8031 at age 6 years and 5560 at age 7 years). Main outcome measures Binary variables of whether a child reached the ‘expected’ level of overall educational achievement across subjects at the ages of 5, 6, 7 and 11 years. The achievement levels are measured using standardised teacher assessments and national tests. Results Compared with full-term births (39–41 weeks), there were significantly increased adjusted odds of children born LPT, but not early term, of failing to achieve expected levels of overall educational achievement at ages 5 years (adjusted OR (aOR) 1.72,95% CI 1.34 to 2.21) and 7 years (aOR 1.46, 95% CI 1.08 to 1.97) but not at age 11 years (aOR 1.51, 95% CI 0.99 to 2.30). Being born LPT still had statistically significant effects on writing and mathematics at age 11 years. Conclusions There is a strong association between LPT and education at age 5 years, which remains strong and statistically significant through age 11 years for mathematics but not for other key subjects.

Refactoring has emerged as a predominant approach to augmenting software product quality. However, empirical evidence suggests that not all dimensions of software quality experience unending enhancements through refactoring. Current scholarly explorations reveal significant variances in the impacts of diverse refactoring methods, with potential adverse effects and contradictions surfacing concerning software quality. Consequently, such disparities render the advantages of refactoring contentious, culminating in challenges for software developers in the selection of optimal refactoring methods to ameliorate software quality. Existing literature lacks an in-depth exploration of the reasons behind the contrasting impacts of refactoring methods on quality enhancement or the development of refined protocols for employing these techniques. Therefore, this research aims to explore, identify, and fine-tune the utilization mechanisms of refactoring methods, empowering software developers to make informed choices for the enhancement of object-oriented systems’ quality attributes. Ten commonly employed refactoring methods were singled out for this investigation, each executed independently across five case studies varying in scale (small, medium, and large). The Quality Model for Object-Oriented Design (QMOOD) was employed as the evaluation tool to ascertain the influence of refactoring techniques on quality attributes. The research outcomes denote that the multifarious impacts of refactoring methods on quality attributes are attributed to distinct usage mechanisms of the techniques. These insights assist software practitioners in discerning the optimal utilization of refactoring methods to ameliorate software quality, taking their mechanisms into account. Moreover, these outcomes furnish industry experts with prescriptive guidelines for employing refactoring methods to elevate the quality of object-oriented systems, predicated on the suitable mechanism.

According to the literature, the success of deinstitutionalization (DI) practices in low-and middle-income countries (LMICs) is dependent on key factors including, a well-functioning family-based alternative care and social protection system, adequate funding and resources, and professional and other stakeholders' engagement and support. Following a practice research qualitative method, the study explored practitioner's experiences and perceptions on the status of Ghana's ongoing DI efforts and their recommendations for improving implementation. The study's main themes were establishing the need for residential homes for children (RHCs), RHCs not being an ideal family environment and RHCs as respite. Family marital problems, poor financial situation, stigma attached to some children in care, abusive parents and a lack of suitable alternatives when families have a crisis were identified as key factors that impede DI implementation in Ghana. The findings suggest the need for a progressive approach towards DI implementation in LMICs, with the first step being the re-positioning of RHCs as respite care centres while progressively developing other alternative family-based care options (such as kinship care) for children. K E Y W O R D S alternative care, deinstitutionalization, family-based care, Ghana, residential homes

Aging is associated with a greater risk of muscle and bone disorders such as sarcopenia and osteoporosis. These conditions substantially affect one’s mobility and quality of life. In the past, muscles and bones are often studied separately using generic or scaled information that are not personal-specific, nor are they representative of the large variations seen in the elderly population. Consequently, the mechanical interaction between the aged muscle and bone is not well understood, especially when carrying out daily activities. This study presents a coupling approach across the body and the organ level, using fully personal-specific musculoskeletal and finite element models in order to study femoral loading during level walking. Variations in lower limb muscle volume/force were examined using a virtual population. These muscle forces were then applied to the finite element model of the femur to study the variations in predicted strains. The study shows that effective coupling across two scales can be carried out to study the muscle-bone interaction in elderly women. The generation of a virtual population is a feasible approach to augment anatomical variations based on a small population that could mimic variations seen in a larger cohort. This is a valuable alternative to overcome the limitation or the need to collect dataset from a large population, which is both time and resource consuming.

The next generation of railway customer-oriented services are expected to generate a large volume of data (~ 10s of TB). As a result, passengers' applications, safety, security, and Internet-on-Board (IoB) sensors challenge current Train Communication Networks. With the present Ethernet Train Backbone (ETB) specification of just 100 Mbit/s in total, railway passenger services will not support intelligent, seamlessly connected and mobile media on-board trains. In this paper, we propose a novel ETB design with experimental results demonstrating end-to-end 10 Gbit/s and 40 Gbit/s throughput results over existing conducting media on commercial railway carriages. This is equivalent to 100 Mbit/s per user on real-world railway rolling stock and shows that standard RailCat 5e cabling and new rail-approved 10 Gbit/s ETB active nodes (switches) fully support emerging trends and future-proof ETB configurations.

After centuries of development, it has been proven that research and development (R&D) and innovation are the driving forces for a country to sustain its economic development and they are also essential to the development of companies. Scientific and technological innovation is strategic support to improve productivity and comprehensive national power, so companies' investment in innovation is increasing year by year. The main output of enterprises is concentrated in the fields of utility models, designs, and applied technology research and development. At the same time, national government policy has placed emphasis on the development of enterprise content and technological innovation, implementing a dual incentive system for enterprise R&D and innovation. This move has helped to energise companies in terms of entrepreneurship and innovation, while also reducing the cost of innovation and driving many new companies to invest in R&D. In this paper, the author discusses whether R&D innovation helps companies to grow by investigating the companys size, governance structure, financial report, and A-share market returns. It is found that R&D innovation is a good driver of business development and can bring advantages to companies.

The envelope model is a useful statistical technique that can be applied to multivariate linear regression problems. It aims to remove immaterial information via sufficient dimension reduction techniques while still gaining efficiency and providing accurate parameter estimates. Recently, envelope tensor versions have been developed to extend this technique to tensor data. In this work, a partial tensor envelope model is proposed that allows for a parsimonious version of tensor response regression when only certain predictors are of interest. The consistency and asymptotic normality of the regression coefficients estimator are also established theoretically, which provides a rigorous foundation for the proposed method. In numerical studies using both simulated and real‐world data, the partial tensor envelope model is shown to outperform several existing methods in terms of the efficiency of the regression coefficients associated with the selected predictors.

Although research on return migration is growing, little is known about returnees’ plans and attitudes regarding further migration. This article contributes to the filling of this knowledge gap by studying the likelihood of engaging in further mobility among Polish and Lithuanian returnees. Using a mixed method approach we explore under which circumstances return migrants intent to stay in their country of origin permanently and what factors would make them consider leaving again. Our quantitative sample (CAWI survey) consists of 740 responses from Poles and Lithuanians who returned to their home countries from the UK. We conducted a binary logistic regression analysis concerning plans to move abroad again. In the qualitative part of the analysis, based on in-depth interviews with 60 Polish or Lithuanian returnees, we have contextualised quantitative results by presenting four case studies representing different likelihoods of re-migrating. Our research shows that both return and post-return plans are always negotiated in the context of a variety of personal, family and professional considerations. Having a job, having children and strong attachment to the current place of living turned out to be the strongest negative predictors of the likelihood of further migration.

This paper investigates whether and how shareholder litigation influences income smoothing. Using the ruling of the Ninth Circuit Court of Appeals in 1999 as an exogenous shock to the threat of litigation, we find that the increasing difficulty of class action lawsuits decreases income smoothing. This finding is robust to different model specifications. We also show that such an effect is stronger for firms that are more likely to face greater pressure from the threat of shareholder litigation risk. Overall, our findings extend the literature on investigating how class action lawsuits can affect the motivation of income smoothing.