MRC Laboratory of Molecular Biology
CUED - Cambridge University Engineering DepartmentNot a member yet
45551 research outputs found
Sort by
Multiphase Engineered BNT-Based Ceramics with Simultaneous High Polarization and Superior Breakdown Strength for Energy Storage Applications
No full textDielectric ceramics are crucial for high-temperature, pulse-power energy storage applications. However, the mutual restriction between the polarization and breakdown strength has been a significant challenge. Here, multiphase engineering controlled by the two-step sintering heating rate is adopted to simultaneously obtain a high polarization and breakdown strength in 0.8(0.95Bi0.5Na0.5TiO3-0.05SrZrO3)-0.2NaNbO3 (BNTSZNN) ceramic systems. The coexistence of tetragonal (T) and rhombohedral (R) phases benefits the temperature stability of BNTSZNN ceramics. Increasing the heating rate during sintering reduces the diffusion of SrZrO3 and NaNbO3 into Bi0.5Na0.5TiO3, which results in a high proportion of the R phase and a finer grain size. The overall polarization is enhanced by increasing the proportion of the high-polarization R phase, which is demonstrated using a first-principles method. Meanwhile, the finer grain size enhances the breakdown strength. Following this design philosophy, an ultrahigh Wdis of 5.55 J/cm3 and I above 85% is achieved in BNTSZNN ceramics as prepared with a fast heating rate of 60 °C/min given a simultaneously high polarization of 43 μC/cm2 and high breakdown strength of 350 kV/cm. Variations in the discharge energy density from room temperature to 160 °C are less than 10%. Additionally, such BNTSZNN ceramics exhibit an ultrafast discharge speed with τ0.9 at approximately 60 ns, which shows great potential in pulse-power system applications
Multitemporal Relearning with Convolutional LSTM Models for Land Use Classification
No full textIn this article, we present a novel hybrid framework, which integrates spatial-Temporal semantic segmentation with postclassification relearning, for multitemporal land use and land cover (LULC) classification based on very high resolution (VHR) satellite imagery. To efficiently obtain optimal multitemporal LULC classification maps, the hybrid framework utilizes a spatial-Temporal semantic segmentation model to harness temporal dependency for extracting high-level spatial-Temporal features. In addition, the principle of postclassification relearning is adopted to efficiently optimize model output. Thereby, the initial outcome of a semantic segmentation model is provided to a subsequent model via an extended input space to guide the learning of discriminative feature representations in an end-To-end fashion. Last, object-based voting is coupled with postclassification relearning for coping with the high intraclass and low interclass variances. The framework was tested with two different postclassification relearning strategies (i.e., pixel-based relearning and object-based relearning) and three convolutional neural network models, i.e., UNet, a simple Convolutional LSTM, and a UNet Convolutional-LSTM. The experiments were conducted on two datasets with LULC labels that contain rich semantic information and variant building morphologic features (e.g., informal settlements). Each dataset contains four time steps from WorldView-2 and Quickbird imagery. The experimental results unambiguously underline that the proposed framework is efficient in terms of classifying complex LULC maps with multitemporal VHR images
Arylethynyltrifluoroborate Dienophiles for on Demand Activation of IEDDA Reactions
No full textStrained alkenes and alkynes are the predominant dienophiles used in inverse electron demand Diels-Alder (IEDDA) reactions. However, their instability, cross-reactivity, and accessibility are problematic. Unstrained dienophiles, although physiologically stable and synthetically accessible, react with tetrazines significantly slower relative to strained variants. Here we report the development of potassium arylethynyltrifluoroborates as unstrained dienophiles for fast, chemically triggered IEDDA reactions. By varying the substituents on the tetrazine (e.g., pyridyl- to benzyl-substituents), cycloaddition kinetics can vary from fast (k2 = 21 M-1 s-1) to no reaction with an alkyne-BF3 dienophile. The reported system was applied to protein labeling both in the test tube and fixed cells and even enabled mutually orthogonal labeling of two distinct proteins
Roadmapping and Roadmaps: Definition and Underpinning Concepts
No full textRoadmapping emerged from industry and has evolved over the decades, through improvements and refinements made by both practitioners and research groups, to become an established and extensively deployed method. Roadmaps are popular in helping to convey and communicate the essence of strategic plans, organizational initiatives, program pathways, and future courses of action. But what actually constitutes a roadmap? What are the unique attributes that distinguish them from other journey-mapping approaches and forward-looking business documents? Drawing upon active involvement in industrial engagements, applied research, tool development, and supported by the literature, a roadmap has now been defined as a structured visual chronology of strategic intent. Further, roadmapping has been defined as the application of a temporal–spatial structured strategic lens. As a result of these more rigorous and robust expressions, this article reports and reviews their underpinning concepts and dimensions, and puts them forward as the new standard definitions
A machine learning approach for predicting hidden links in supply chain with graph neural networks
No full textSupply chain business interruption has been identified as a key risk factor in recent years, with high-impact disruptions due to disease outbreaks, logistic issues such as the recent Suez Canal blockage showing examples of how disruptions could propagate across complex emergent networks. Researchers have highlighted the importance of gaining visibility into procurement interdependencies between suppliers to develop more informed business contingency plans. However, extant methods such as supplier surveys rely on the willingness or ability of suppliers to share data and are not easily verifiable. In this article, we pose the supply chain visibility problem as a link prediction problem from the field of Machine Learning (ML) and propose the use of an automated method to detect potential links that are unknown to the buyer with Graph Neural Networks (GNN). Using a real automotive network as a test case, we show that our method performs better than existing algorithms. Additionally, we use Integrated Gradient to improve the explainability of our approach by highlighting input features that influence GNN’s decisions. We also discuss the advantages and limitations of using GNN for link prediction, outlining future research directions
Laminar flow-induced scission kinetics of polymers in dilute solutions
No full textMechanical degradation of macromolecules in strong flows is encountered in many industrial processes spanning from biopharmaceutics manufacturing to enhanced oil recovery. In spite of extensive research, from molecular studies to large experiments, unifying scaling laws and design rules to harness this phenomenon are still at an early stage. Some of the current modelling approaches predict the onset of flow-induced degradation only, leaving out quantitative calculations of scission events, while others are restricted to a particular process or the materials they have been empirically developed for. In this work we re-examine a previously published constitutive equation for the scission kinetics of polymers and implement the model using the finite volume library OpenFoam. We test and validate this model using experimental degradation measurements of aqueous poly(ethylene oxide) solutions flowing through narrow constrictions. Three polymer molecular weights and three constriction geometries are investigated. For each molecular weight, experimental degradation data of one geometry is used to calibrate the model. Following this calibration step, the level of polymer degradation as a function of flow rate can be predicted for the two other geometries, suggesting that mechanisms linking single molecule scission to macroscopic chemical reaction rate are accurately captured by the model. Although the focus of this work is on flexible linear polymers in dilute concentrations and laminar flow conditions, we discuss how to alleviate these assumptions and extend the applicability of the model to a broader range of materials and industrially relevant flow conditions
Experimental modelling of seasonal thermal energy storage within unconfined aquifer (Ates)
No full textAquifer thermal energy storage systems allow the storage of excess heat from summer for use during the winter. This investigation looks at the suitability of a small-scale experimental model as a method for simulating the behaviour of full-scale unconfined aquifers for thermal storage. Thermal energy was stored via the injection of 40, 60, and 80 °C water for a period of 1000 s with extraction being between 1000 and 2000 s. Furthermore, periods of storage between injection and extraction were introduced to simulate potential full-scale heating and cooling demand scenarios. Thermal efficiencies were found to be 60% reducing to 53% with the addition of a 1000 s storage period. Furthermore, for the model tested in this investigation the temperature of the injected water was found to have little influence upon the efficiency
Application of machine learning for filtered density function closure in MILD combustion
No full textA machine learning algorithm, the deep neural network (DNN)1, is trained using a comprehensive direct numerical simulation (DNS) dataset to predict joint filtered density functions (FDFs) of mixture fraction and reaction progress variable in Moderate or Intense Low-oxygen Dilution (MILD) combustion. The important features of the DNS cases include mixture fraction variations, turbulent mixing lengths, exhaust gas recirculation (EGR) dilution levels, etc., posing a great challenge for data-driven modelling. The DNN architecture is built and optimised with extreme care to achieve high robustness and accuracy, resorting to dimensionality reduction techniques such as principal component analysis (PCA) to identify and remove the outliers in the training data. To better interpret the predictive ability of the DNN, two analytical joint FDF models respectively using two independent β and copula distributions, are also employed for a detailed comparison with the DNS data. The FDFs in MILD combustion behave differently compared to those in conventional flames because the reaction zones are more distributed. They generally exhibit non-regular (neither Gaussian nor bi-modal) distributions and strong cross correlations, which cannot be captured adequately by the analytical models. However, the DNN is well suited for this physico-chemically complex problem and its predictions are in excellent agreement with the DNS data for a broad range of mixture conditions and filter sizes. Furthermore, a priori assessment is conducted for filtered reaction rate closure. It is found that the DNN model significantly outperforms the analytical models for all cases showing very good predictions for the filtered reaction rate for a range of filter sizes. The DNN prediction improves as the filter size becomes larger than the characteristic reaction zone thickness while the analytical models works relatively better for smaller filter sizes. This is a clear advantage for the DNN to be used in practical LES applications
Molecular dynamics investigation on the interfacial shear creep between carbon fiber and epoxy matrix
No full textCarbon fiber-reinforced polymer (CFRP) composite is subject to external loads during service life, suffering the interfacial creep between the fiber and the matrix and eventually the interfacial slippage. CFRP composite loses the interfacial integrity due to the interfacial creep and the long-term durability is weakened. In order to understand the degradation, microscopic details of interfacial structural changes during the creep are essential. This study aims to investigate microscopic creep behavior of a carbon fiber/epoxy interface at different shear load levels using molecular dynamics simulations. A molecular interface model consisting of an epoxy molecule bonded to graphite sheets representing the fiber outer layer is constructed, which is validated by comparing the mass density, glass-transition temperature and Young's modulus of bonded epoxy with experimental measurements. According to the creep simulation, there is a threshold stress for the onset of creep failure, above which the interface detaches. Comparatively, no interfacial detachment occurs under the low stress regime, where the displacement–force curve is plotted and used to quantify the energy barrier to the onset of creep failure. Meanwhile, the strain and stress evolution of the interface are correlated to interfacial structural changes to understand the interfacial creep mechanism. This study provides molecular insights into the interfacial creep behavior in the fiber/matrix system and form the basis of multiscale investigation framework on the interfacial creep behavior
Uncertainty as a Form of Transparency: Measuring, Communicating, and Using Uncertainty
No full textAlgorithmic transparency entails exposing system properties to various stakeholders for purposes that include understanding, improving, and contesting predictions. Until now, most research into algorithmic transparency has predominantly focused on explainability. Explainability attempts to provide reasons for a machine learning model's behavior to stakeholders. However, understanding a model's specific behavior alone might not be enough for stakeholders to gauge whether the model is wrong or lacks sufficient knowledge to solve the task at hand. In this paper, we argue for considering a complementary form of transparency by estimating and communicating the uncertainty associated with model predictions. First, we discuss methods for assessing uncertainty. Then, we characterize how uncertainty can be used to mitigate model unfairness, augment decision-making, and build trustworthy systems. Finally, we outline methods for displaying uncertainty to stakeholders and recommend how to collect information required for incorporating uncertainty into existing ML pipelines. This work constitutes an interdisciplinary review drawn from literature spanning machine learning, visualization/HCI, design, decision-making, and fairness. We aim to encourage researchers and practitioners to measure, communicate, and use uncertainty as a form of transparency