Search

This work presents a modular Smart Maintenance Platform (SMP), which offers a toolbox of condition monitoring tools for the day-to-day inspection of the machinery equipment in a real industrial environment. SMP complements the manual inspection performed by the maintenance engineers with advanced data-analytics solutions that utilize both reactive and proactive techniques based on state-of-the-art machine learning proposals. Appropriate data acquisition mechanisms are built to acquire the sensorial data and transform them to suitable forms for data processing. Higher-level monitoring policies are enabled, through the fusion of multiple monitoring tasks increasing the precision of the results. The generated results are persisted for the continuous enhancement of the solution through the retraining of the proactive approach or the re-configuration of the reactive ones. The generated results are visualized in a user-friendly format, to help the maintenance engineers to assess the severity of the situation and to intervene in the machinery equipment if deemed necessary. [ABSTRACT FROM AUTHOR]

In the last decades, major advances in the language sciences have been built on real-time measures of language and cognitive processing, measures like mouse-tracking, event related potentials and eye-tracking in the visual world paradigm. These measures yield densely sampled timeseries that can be highly revealing of the dynamics of cognitive processing. However, despite these methodological advances, existing statistical approaches for timeseries analyses have often lagged behind. Here, we present a new statistical approach, the Bootstrapped Differences of Timeseries (BDOTS), that can estimate the precise timewindow at which two timeseries differ. BDOTS makes minimal assumptions about the error distribution, uses a custom family-wise error correction, and can flexibly be adapted to a variety of applications. This manuscript presents the theoretical basis of this approach, describes implementational issues (in the associated R package), and illustrates this technique with an analysis of an existing dataset. Pitfalls and hazards are also discussed, along with suggestions for reporting in the literature. [ABSTRACT FROM AUTHOR]

While the concept of connectivity has gained popularity in fields like hydrology and ecology, little agreement exists on its definition, which hinders its use in both scientific and legal contexts. In contrast, neuroscientists have developed not only strong conceptualizations of connectivity but also tools to quantify it: a clear distinction is made between structural connectivity, which is determined from brain anatomy; functional connectivity, which is estimated based on statistical dependencies between neuronal electric timeseries; and effective connectivity, which infers causal relations from the same timeseries based on the assumption that “true” interactions occur with a certain time delay. The motivation of this review arose from the hypothesis that connectivity-related statistical techniques, which are applied to timeseries of electrical currents measured by placing electrodes on the scalp of the human brain, could also apply to high-frequency hydrological timeseries acquired to characterize catchment response to precipitation. Here we bring together existing conceptualizations of structural, functional and effective connectivity in hydrology and ecology and compare them with those used in brain neuroscience. We then summarize the most important brain connectivity measures and their associated mathematical frameworks before evaluating the potential of those measures to help advance our understanding of hydrologic connectivity properties – in terms of the frequency, magnitude, timing, duration and rate of water movement linking two disparate locations. Lastly, we present a short case study where a selection of brain connectivity measures is applied to 35 groundwater and streamflow timeseries from a Swiss catchment to infer subsurface flow-driven hydrologic connectivity. Our literature review combined with our short case study suggest that an ensemble of functional and effective connectivity measures should be used and constrained not only by structural connectivity measures but also by interpretation thresholds in order to make results parsimonious. We highlight challenges associated with transferring brain connectivity measures to hydrology, especially those related to choosing the appropriate length and sampling frequency of input timeseries when assessing perennial versus ephemeral connectivity, appropriately detecting and differentiating noisy from indirect connections, and interpreting unbounded connectivity measures. We then offer recommendations for future research and propose that hydrologists use a common classification system encompassing all potential connectivity assessment approaches and measures in order to facilitate scientific communication. [ABSTRACT FROM AUTHOR]

Student cooperativeness underlies high quality teacher-student relationships, and has been positively associated with students' school engagement. Fostering cooperative rather than oppositional student behavior might be especially helpful for protecting at-risk students against academic failure. To understand how exactly students' cooperativeness can be fostered, we investigated the interpersonal behaviors of secondary school teachers and at-risk students during dyadic interactions (N = 82 dyads) in the context of positive teacher-student relationships. Using Continuous Assessment of Interpersonal Dynamics, moment-to-moment teacher and student behavior was captured in terms of interpersonal agency (dominance vs. submissiveness) and communion (opposition vs. cooperation). Time-series analyses were used to analyze interpersonal behavior within individuals, within dyads, and between dyads. Cooperative student behavior was most likely if teachers acted friendly and cooperatively and if teachers 'loosened up' their agency and the structure they imposed on the interaction repeatedly, which may give students more freedom to express themselves and to cooperate. • Dyadic interaction between at-risk students and their favorite teacher was assessed. • Observations captured both agency and communion of teachers and students. • Student cooperation was fostered by teacher cooperation. • Students were more cooperative when teachers showed flexibility in agency. • When fostering student cooperation both communion and agency are important. [ABSTRACT FROM AUTHOR]

Abstract: Cyclical patterns of behaviour in timeseries of seismic and geodetic data at volcanoes are frequently observed during lava dome-building eruptions, and are particularly well-documented from the current eruption of the Soufrière Hills Volcano (SHV), Montserrat. However, the discontinuous nature of many SO2 measurements often preclude the identification and quantitative analysis of cyclical patterns in degassing data. Here, using a long SO2 timeseries from SHV, with continuous measurements since 2002, we explore for the first time degassing behaviour at a resolution comparable to that possible for seismic and deformation datasets. Timeseries analysis of flux data spanning 2002–2011 reveals that SO2 emissions at SHV exhibit complex cyclicity, with dominant cycles evident on both multi-year and multi-week (~50 day) timescales. These cycles persist through phases of both active extrusion and eruptive pause, and show close similarities to periodic components previously identified at SHV in timeseries of seismicity, ground deformation and lava extrusion. The strength of expression or amplitude of degassing cycles, particularly on multi-week timescales, shows distinct temporal variation, and appears to correlate with the occurrence and nature of explosive activity occurring in 2002–2009. This suggests that the amplitude of surface gas flux cycles is modulated by physical conditions within the conduit. Direct quantitative comparison between seismicity, dome growth, and degassing for eruptive Phases 2 (2002–2003) and 3 (2005–2007) reveals that peaks in SO2 flux appear to correspond broadly to enhanced lava extrusion and elevated seismicity within cycles of 30–50 days. However, time lags of 2, 4 and 7 days are observed between initial low-frequency seismic swarms and peaks in dome growth, SO2 flux and rockfall event rate respectively. Multi-parameter correlations offer valuable insights into the controls on subsurface gas ascent, but further research is required to fully explore the contributions of permeability and overpressure, as well as other subsurface processes. [Copyright &y& Elsevier]

Abstract: We explore the use of block entropy as a dynamics classifier for meteorological timeseries data. The block entropy estimates define the entropy growth curve H(L) with respect to block length L. For a finitary process, the entropy growth curve tends to an asymptotic linear regime H(L) = E + hμL, with entropy rate hμ and excess entropy E. These quantities apportion the system''s information content into ‘memory’ (E) and ‘randomness’ (hμ). We discuss the challenges inherent in analyzing weather data using symbolic techniques, identifying the pitfalls associated with alphabet size, finite sample timeseries length, and stationarity. We apply the block entropy-based techniques in the form of a wet/dry partition to Australian daily precipitation data from the Patched Point Dataset station record collection and version 3 of the Australian Water Availability Project analysis dataset. Preliminary results demonstrate hμ and E are viable climatological classifiers for precipitation, with station records from similar climatic regimes possessing similar values of hμ and E. The resultant clustering reflects expected characteristics of local climatic memory and randomness. The AWAP results show weaker clustering than their PPD counterparts, with different E- and hμ-values reflecting respectively the relative biases and truncation errors in the AWAP analysis system. The entropy rates of convergence analysis rules out finite order Markov processes for orders falling within the range of block sizes considered. [Copyright &y& Elsevier]

• We developed a Budyko-based approach for detection of fire effects on streamflow. • We attributed changes in streamflow in post-fire years to fire and climatic causes. • Hydrologic recovery ranged between 5 and 45 years, increasing with increased burned area. Recent increases in the incidences of wildfires have necessitated the development of methodologies to quantify the effect of these fires on streamflows. Climate variability has been cited as a major challenge in revealing the true contribution of disturbance to streamflow changes. To address this, we developed an annual Budyko "decomposition" method for (1) statistical change detection of hydrologic signatures post-fire, (2) separating climate-driven and fire-driven changes in streamflow, and (3) estimating hydrologic recovery timescales after fire. We demonstrate the use of this methodology for 17 watersheds in Southern California with high interannual variability in precipitation. We show that while traditional metrics like changes in flow or runoff ratio might not detect a disturbance effect due to confounding climate signals, the Budyko framework can be used successfully for statistical change detection. The Budyko approach was also found to be robust in detecting changes in 5 highly burned catchments (>40% burned area ratio), while changes in less burned (2) and unburned catchments (10) were insignificant. We further used the Budyko approach to quantify the contribution of fire-driven versus climate driven changes in streamflow and found that fire contributed to an average increase in streamflow on the order of 80 mm yr−1, though the effect varied greatly between years. Finally, we estimated hydrologic recovery timescales that varied between 5 and 45 years for four burned catchments. We found a significant linear relationship between recovery time and burned area at medium and high severity for our study catchments, with about 4 years of recovery time per 10% of the watershed burned. [ABSTRACT FROM AUTHOR]

We study the temporal variability of human brain activity in timeseries of functional magnetic resonance imaging (fMRI) data. We find that these timeseries show scaling behavior which is quantified by computing various scaling exponents. We demonstrate that mentally active zones are one-to-one related to large exponents, or equivalently, to highly temporally correlated processes, while mentally inactive zones are well described by a simple random walk model. Activation maps of scaling exponents are presented and compared to standard results in fMRI. In contrast to standard model-based activation analyses no prior knowledge of the experimental stimulation paradigm has to be assumed for extracting activation patterns from fMRI timeseries. We mention consequences of persistence in fMRI timeseries for standard statistical analysis. [Copyright &y& Elsevier]

We analyzed the timeseries of power loads of 35 separate countries which publicly share hourly data through the ENTSO-E platform for more than 5 years. Moreover we applied Multifractal Detrended Fluctuation Analysis for the demonstration of the multifractal nature, autocorrelation and the distribution function fundamentals. Additionally, we improved the basic method described by Kantelhardt, et al using uniform shuffling and surrogation of the datasets in order to prove the robustness of the results with respect to the non-linear effects of the processes. All the datasets exhibited multifractality in the distribution function as well as in the autocorrelation function. The basic differences between individual states are manifested in the width of the multifractal spectra and in the location of the maxima. We present the hypothesis concerning the production portfolio and the export/import dependencies. • We observed the 1/f noise property in electricity power loads in Europe. • Multifractality of correlation multifractality and distribution function. • Robustness against non-linear effects. • Classification of results using machine learning algorithm. [ABSTRACT FROM AUTHOR]