Search

A population-level analysis is proposed to address data sparsity when building predictive models for engineering infrastructure. Utilising an interpretable hierarchical Bayesian approach and operational fleet data, domain expertise is naturally encoded (and appropriately shared) between different sub-groups, representing (i) use-type, (ii) component, or (iii) operating condition. Specifically, domain expertise is exploited to constrain the model via assumptions (and prior distributions) allowing the methodology to automatically share information between similar assets, improving the survival analysis of a truck fleet and power prediction in a wind farm. In each asset management example, a set of correlated functions is learnt over the fleet, in a combined inference, to learn a population model. Parameter estimation is improved when sub-fleets share correlated information at different levels of the hierarchy. In turn, groups with incomplete data automatically borrow statistical strength from those that are data-rich. The statistical correlations enable knowledge transfer via Bayesian transfer learning, and the correlations can be inspected to inform which assets share information for which effect (i.e. parameter). Both case studies demonstrate the wide applicability to practical infrastructure monitoring, since the approach is naturally adapted between interpretable fleet models of different in situ examples.

The widespread diffusion of precision radiotherapy techniques, geared toward the release of larger dose gradients in shorter time frames, is leading to new challenges in dosimetry. Accurate dose measurements are essential to check for beam anomalies and inaccuracies to ensure treatment efficacy and patient safety during radiotherapy. This work describes the main features of a diamond dosimeter coupled to an extremely compact front-end electronics. The detection system was tested under the X-ray pulses generated by a medical LINAC for both the 6 MV and the 18 MV accelerating voltages. Located in the LINAC's bunker, it eliminates the need for a long cable connection between the detector and the electronics, detrimental for the system response speed. Signal acquisition was performed synchronously with the impinging X-ray pulses with a sampling period as low as 20 us, allowing for a real-time beam monitoring. The dosimeter demonstrated a very good stability despite the high value of the absorbed dose during the performed experiments (about 100 Gy). The measured dose-per-pulse values of 278 uGy and 556 uGy at 6 MV and 18 MV, respectively, are in excellent agreement with the nominal values expected for the LINAC apparatus used for the tests. In addition to single-pulse measurements, fundamental for dynamic radiotherapy, the proposed system also allows for the calculation of both the total collected charge and the photocurrent generated by the detector. In this regards, despite the compactness, it demonstrates its effectiveness as a tool for source diagnostics in terms of both beam intensity and emission timing., Comment: 14 pages, 8 figures, Prepared for submission to IEEE Sensors Journal

This paper presents a lightweight, open-source and high-performance python package for solving peridynamics problems in solid mechanics. The development of this solver is motivated by the need for fast analysis tools to achieve the large number of simulations required for `outer-loop' applications, including sensitivity analysis, uncertainty quantification and optimisation. Our python software toolbox utilises the heterogeneous nature of OpenCL so that it can be executed on any platform with CPU or GPU cores. We illustrate the package use through a range of industrially motivated examples, which should enable other researchers to build on and extend the solver for use in their own applications. Step improvements in execution speed and functionality over existing techniques are presented. A comparison between this solver and an existing OpenCL implementation in the literature is presented, tested on benchmarks with hundreds of thousands to tens of millions of nodes. We demonstrate the scalability of the solver on the GeForce RTX 2080 TiGPU from NVIDIA, and the memory-bound limitations are analysed. In all test cases, the implementation is between 1.4 and 10.0 times faster than a similar existing GPU implementation in the literature. In particular, this improvement has been achieved by utilising local memory on the GPU., Comment: peripy.readthedocs.org

In the present article we describe the calibration measurements of the compact neutron spectrometer at certified PTB neutron source. The spectrometer is based on two diamond detectors enclosing the Li-6 neutron converter. The incident neutron energy is measured through the sum of energies of the two conversion reaction products: t and alpha. The prototype used in this work was based on two 4x4 mm2, 100 um thick single crystal CVD commercial diamond plates. The measurements were performed in the fission spectrum domain at five different neutron energies varying from 0.3 up to 2 MeV. The obtained data showed a very good neutron energy reconstruction, i.e. in agreement with its reference values within statistical uncertainties or with deviation below 11 keV in absolute scale. The energy resolution of the spectrometer was found to be about 100 keV (RMS) with a small (about 9 keV/MeV) rising trend towards higher energies. The measured detection efficiency was found to be compatible with expectations based on analytical calculations and on Geant4 simulations within 3% statistical and 4% systematic precision., Comment: 20 pages, 14 figures

The building of mathematical and computer models of cities has a long history. The core elements are models of flows (spatial interaction) and the dynamics of structural evolution. In this article, we develop a stochastic model of urban structure to formally account for uncertainty arising from less predictable events. Standard practice has been to calibrate the spatial interaction models independently and to explore the dynamics through simulation. We present two significant results that will be transformative for both elements. First, we represent the structural variables through a single potential function and develop stochastic differential equations (SDEs) to model the evolution. Secondly, we show that the parameters of the spatial interaction model can be estimated from the structure alone, independently of flow data, using the Bayesian inferential framework. The posterior distribution is doubly intractable and poses significant computational challenges that we overcome using Markov chain Monte Carlo (MCMC) methods. We demonstrate our methodology with a case study on the London retail system., Comment: http://dx.doi.org/10.1098/rspa.2017.0700

In this paper a new version of Li6-based neutron spectrometer for high flux environments is described. The new spectrometer was built with commercial single crystal Chemical Vapour Deposition diamonds of electronic grade. These crystals feature better charge collection as well as higher radiation hardness. Ohmic metal contacts were deposited on the diamonds suppressing build-up of space charge observed in the previous prototypes. New passive preamplification of signal at detector side was implemented to improve the resolution. This preamplification is based on RF transformer not sensitive to high neutron flux. Compact mechanical design allowed to reduce detector size to a tube of 1 cm diameter and 13 cm long. The spectrometer was tested in thermal column of TRIGA reactor and at DD neutron generator. The test results indicate an energy resolution of 72 keV (RMS) and coincidence timing resolution of 68 ps (RMS). The measured data are in agreement with Geant4 simulations except for larger energy loss tail presumably related to imperfections of metal contacts and glue expansion., Comment: 18 pages, 10 figures

For many neurological disorders, prediction of disease state is an important clinical aim. Neuroimaging provides detailed information about brain structure and function from which such predictions may be statistically derived. A multinomial logit model with Gaussian process priors is proposed to: (i) predict disease state based on whole-brain neuroimaging data and (ii) analyze the relative informativeness of different image modalities and brain regions. Advanced Markov chain Monte Carlo methods are employed to perform posterior inference over the model. This paper reports a statistical assessment of multiple neuroimaging modalities applied to the discrimination of three Parkinsonian neurological disorders from one another and healthy controls, showing promising predictive performance of disease states when compared to nonprobabilistic classifiers based on multiple modalities. The statistical analysis also quantifies the relative importance of different neuroimaging measures and brain regions in discriminating between these diseases and suggests that for prediction there is little benefit in acquiring multiple neuroimaging sequences. Finally, the predictive capability of different brain regions is found to be in accordance with the regional pathology of the diseases as reported in the clinical literature., Comment: Published in at http://dx.doi.org/10.1214/12-AOAS562 the Annals of Applied Statistics (http://www.imstat.org/aoas/) by the Institute of Mathematical Statistics (http://www.imstat.org)