Active learning strategies for robotic tactile texture recognition tasks.

Accurate texture classification empowers robots to improve their perception and comprehension of the environment, enabling informed decision-making and appropriate responses to diverse materials and surfaces. Still, there are challenges for texture classification regarding the vast amount of time series data generated from robots' sensors. For instance, robots are anticipated to leverage human feedback during interactions with the environment, particularly in cases of misclassification or uncertainty. With the diversity of objects and textures in daily activities, Active Learning (AL) can be employed to minimize the number of samples the robot needs to request from humans, streamlining the learning process. In the present work, we use AL to select the most informative samples for annotation, thus reducing the human labeling effort required to achieve high performance for classifying textures. We also use a sliding window strategy for extracting features from the sensor's time series used in our experiments. Our multi-class dataset (e.g., 12 textures) challenges traditional AL strategies since standard techniques cannot control the number of instances per class selected to be labeled. Therefore, we propose a novel class-balancing instance selection algorithm that we integrate with standard AL strategies. Moreover, we evaluate the effect of sliding windows of two-time intervals (3 and 6 s) on our AL Strategies. Finally, we analyze in our experiments the performance of AL strategies, with and without the balancing algorithm, regarding f1-score, and positive effects are observed in terms of performance when using our proposed data pipeline. Our results show that the training data can be reduced to 70% using an AL strategy regardless of the machine learning model and reach, and in many cases, surpass a baseline performance. Finally, exploring the textures with a 6-s window achieves the best performance, and using either Extra Trees produces an average f1-score of 90.21% in the texture classification data set.

A temporal assessment of risk of non-indigenous species introduction by ballast water to Canadian coastal waters based on environmental similarity.

The environmental similarity scores between source and recipient locations are essential in ballast water risk assessment (BWRA) models used to estimate the potential for non-indigenous species (NIS) introduction, survival, and establishment, and to guide management strategies aiming to minimize biodiversity loss and economic impacts. Previous BWRA models incorporate annual-scale environmental data, which may overlook seasonal variability. In this study, temporal variation in sea surface temperature and salinity data were examined at global ports, and the influence of this variation on environmental distance calculations (and corresponding risk of NIS) was examined for ballast water discharges in Canada by comparing outputs from monthly and annual scale assessments in a BWRA model. Except for some outliers in the Pacific region, the environmental distances based on monthly scale data generally become smaller in all regions, demonstrating that the model using annual decadal average environmental data to inform environmental matching can underestimate risk of NIS survival and establishment in comparison to monthly data. The results of this study suggest future evaluations incorporating the date of ballast water uptake and discharge can provide a more sensitive assessment of risk reflecting seasonal variability compared to an annual average risk model.

A Semi-Supervised Methodology for Fishing Activity Detection Using the Geometry behind the Trajectory of Multiple Vessels.

Automatic Identification System (AIS) messages are useful for tracking vessel activity across oceans worldwide using radio links and satellite transceivers. Such data play a significant role in tracking vessel activity and mapping mobility patterns such as those found during fishing activities. Accordingly, this paper proposes a geometric-driven semi-supervised approach for fishing activity detection from AIS data. Through the proposed methodology, it is shown how to explore the information included in the messages to extract features describing the geometry of the vessel route. To this end, we leverage the unsupervised nature of cluster analysis to label the trajectory geometry, highlighting changes in the vessel's moving pattern, which tends to indicate fishing activity. The labels obtained by the proposed unsupervised approach are used to detect fishing activities, which we approach as a time-series classification task. We propose a solution using recurrent neural networks on AIS data streams with roughly 87% of the overall F-score on the whole trajectories of 50 different unseen fishing vessels. Such results are accompanied by a broad benchmark study assessing the performance of different Recurrent Neural Network (RNN) architectures. In conclusion, this work contributes by proposing a thorough process that includes data preparation, labeling, data modeling, and model validation. Therefore, we present a novel solution for mobility pattern detection that relies upon unfolding the geometry observed in the trajectory.

A Study on the Geometric and Kinematic Descriptors of Trajectories in the Classification of Ship Types.

The classification of ships based on their trajectory descriptors is a common practice that is helpful in various contexts, such as maritime security and traffic management. For the most part, the descriptors are either geometric, which capture the shape of a ship's trajectory, or kinematic, which capture the motion properties of a ship's movement. Understanding the implications of the type of descriptor that is used in classification is important for feature engineering and model interpretation. However, this matter has not yet been deeply studied. This article contributes to feature engineering within this field by introducing proper similarity measures between the descriptors and defining sound benchmark classifiers, based on which we compared the predictive performance of geometric and kinematic descriptors. The performance profiles of geometric and kinematic descriptors, along with several standard tools in interpretable machine learning, helped us to provide an account of how different ships differ in movement. Our results indicated that the predictive performance of geometric and kinematic descriptors varied greatly, depending on the classification problem at hand. We also showed that the movement of certain ship classes solely differed geometrically while some other classes differed kinematically and that this difference could be formulated in simple terms. On the other hand, the movement characteristics of some other ship classes could not be delineated along these lines and were more complicated to express. Finally, this study verified the conjecture that the geometric-kinematic taxonomy could be further developed as a tool for more accessible feature selection.

An urban energy balance-guided machine learning approach for synthetic nocturnal surface Urban Heat Island prediction: A heatwave event in Naples.

Southern European functional urban areas (FUAs) are increasingly subject to heatwave (HW) events, calling for anticipated climate adaptation measures. In the urban context, such adaptation strategies require a thorough understanding of the built-up response to the incoming solar radiation, i.e., the urban energy balance cycle and its implications for the Urban Heat Island (UHI) effect. Despite readily available, diurnal Land Surface Temperature (LST) data does not provide a meaningful picture of the UHI, in these midlatitudes FUAs. On the contrary, the mid-morning satellite overpass is characterized by the absence of a significant surface UHI (SUHI) signal, corresponding to the period of the day when the urban-rural air temperature difference is typically negative. Conversely, nocturnal high-resolution LST data is rarely available. In this study, an energy balance-based machine learning approach is explored, considering the Local Climate Zones (LCZ), to describe the daily cycle of the heat flux components and predict the nocturnal SUHI, during an HW event. While the urban and rural spatial outlines are not visible in the diurnal thermal image, they become apparent in the latent and storage heat flux maps - built-up infrastructures uptake heat during the day which is released back into the atmosphere, during the night, whereas vegetation land surfaces loose diurnal heat through evapotranspiration. For the LST prediction model, a random forest (RF) approach is implemented. RF results show that the model accurately predicts the LST, ensuring mean square errors inferior to 0.1 K. Both the latent and storage heat flux components, together with LCZ classification, are the most important explanatory variables for the nocturnal LST prediction, supporting the adoption of the energy balance approach. In future research, other locations and time-series data shall be trained and tested, providing an efficient local urban climate monitoring tool, where in-situ air temperature observations are not available.

Examining the impact of cross-domain learning on crime prediction.

Nowadays, urban data such as demographics, infrastructure, and criminal records are becoming more accessible to researchers. This has led to improvements in quantitative crime research for predicting future crime occurrence by identifying factors and knowledge from instances that contribute to criminal activities. While crime distribution in the geographic space is asymmetric, there are often analog, implicit criminogenic factors hidden in the data. And, since the data are not as available or comprehensive, especially for smaller cities, it is challenging to build a uniform framework for all geographic regions. This paper addresses the crime prediction task from a cross-domain perspective to tackle the data insufficiency problem in a small city. We create a uniform outline for Halifax, Nova Scotia, one of Canada's geographic regions, by adapting and learning knowledge from two different domains, Toronto and Vancouver, which belong to different but related distributions with Halifax. For transferring knowledge among source and target domains, we propose applying instance-based transfer learning settings. Each setting is directed to learning knowledge based on a seasonal perspective with cross-domain data fusion. We choose ensemble learning methods for model building as it has generalization capabilities over new data. We evaluate the classification performance for both single and multi-domain representations and compare the results with baseline models. Our findings exhibit the satisfactory performance of our proposed data-driven approach by integrating multiple sources of data.

Aircraft Fuselage Corrosion Detection Using Artificial Intelligence.

Corrosion identification and repair is a vital task in aircraft maintenance to ensure continued structural integrity. Regarding fuselage lap joints, typically, visual inspections are followed by non-destructive methodologies, which are time-consuming. The visual inspection of large areas suffers not only from subjectivity but also from the variable probability of corrosion detection, which is aggravated by the multiple layers used in fuselage construction. In this paper, we propose a methodology for automatic image-based corrosion detection of aircraft structures using deep neural networks. For machine learning, we use a dataset that consists of D-Sight Aircraft Inspection System (DAIS) images from different lap joints of Boeing and Airbus aircrafts. We also employ transfer learning to overcome the shortage of aircraft corrosion images. With precision of over 93%, we demonstrate that our approach detects corrosion with a precision comparable to that of trained operators, aiding to reduce the uncertainties related to operator fatigue or inadequate training. Our results indicate that our methodology can support specialists and engineers in corrosion monitoring in the aerospace industry, potentially contributing to the automation of condition-based maintenance protocols.

An urban climate-based empirical model to predict present and future patterns of the Urban Thermal Signal.

Air temperature is a key aspect of urban environmental health, especially considering population and climate change prospects. While the urban heat island (UHI) effect may aggravate thermal exposure, city-level UHI regression studies are generally restricted to temporal-aggregated intensities (e.g., seasonal), as a function of time-fixed factors (e.g., urban density). Hence, such approaches do not disclose daily urban-rural air temperature changes, such as during heatwaves (HW). Here, summer data from Lisbon's air temperature urban network (June to September 2005-2014), is used to develop a linear mixed-effects model (LMM) to predict the daily median and maximum Urban Thermal Signal (UTS) intensities, as a response to the interactions between the time-varying background weather variables (i.e., the regional/non-urban air temperature, 2-hours air temperature change, and wind speed), and time-fixed urban and geographic factors (local climate zones and directional topographic exposure). Results show that, in Lisbon, greatest temperatures and UTS intensities are found in 'Compact' areas of the city are proportional to the background air temperature change. In leeward locations, the UTS can be enhanced by the topographic shelter effect, depending on wind speed - i.e., as wind speed augments, the UTS intensity increases in leeward sites, even where sparsely built. The UTS response to a future urban densification scenario, considering climate change HW conditions (RCP8.5, 2081-2100 period), was also assessed, its results showing an UTS increase of circa 1.0 °C, in critical areas of the city, despite their upwind location. This LMM empirical approach provides a straightforward tool for local authorities to: (i) identify the short-term critical areas of the city, to prioritise public health measures, especially during HW events; and (ii) test the urban thermal performance, in response to climate change and urban planning scenarios. While the model coefficient estimates are case-specific, the approach can be efficiently replicated in other locations with similar biogeographic conditions.

Geostatistical COVID-19 infection risk maps for Portugal.

The rapid spread of the SARS-CoV-2 epidemic has simultaneous time and space dynamics. This behaviour results from a complex combination of factors, including social ones, which lead to significant differences in the evolution of the spatiotemporal pattern between and within countries. Usually, spatial smoothing techniques are used to map health outcomes, and rarely uncertainty of the spatial predictions are assessed. As an alternative, we propose to apply direct block sequential simulation to model the spatial distribution of the COVID-19 infection risk in mainland Portugal. Given the daily number of infection data provided by the Portuguese Directorate-General for Health, the daily updates of infection rates are calculated by municipality and used as experimental data in the geostatistical simulation. The model considers the uncertainty/error associated with the size of each municipality's population. The calculation of daily updates of the infection risk maps results from the median model of one ensemble of 100 geostatistical realizations of daily updates of the infection risk. The ensemble of geostatistical realizations is also used to calculate the associated spatial uncertainty of the spatial prediction using the interquartile distance. The risk maps are updated daily and show the regions with greater risks of infection and the critical dynamics related to its development over time.

Assessing human exposure to polycyclic aromatic hydrocarbons (PAH) in a petrochemical region utilizing data from environmental biomonitors.

Polycyclic aromatic hydrocarbons (PAH) are toxic compounds that have been classified by the International Agency for Research on Cancer as probable or possible human carcinogens. Human exposure to PAH is usually assessed by considering data from a single air monitoring station as being representative of a large region; however, air pollution levels change on small spatial scales and thus also affect environmental exposure. The use of environmental biomonitors is a useful tool to assess the levels of PAH with high spatial resolution. The aims of this study were to (1) assess human exposure to PAH in a petrochemical region in Portugal, integrating data from environmental biomonitors (lichens), air, and soil in a regional area, and (2) determine the health risks associated with exposure to PAH with high spatial resolution. Bearing this in mind, benzo[a]pyrene (BaP) equivalent concentrations in samples of soil, air, and lichens collected in the study region were used to assess human exposure through different pathways, including inhalation of air and soil particles, ingestion of soil, and dermal contact with soil. Human health risk was calculated through the Incremental Lifetime Cancer Risk (ILCR). BaP equivalent concentrations found in the region ranged from 6.9 to 46.05 ng BaPeq/g in lichens, from 16.45 to 162.02 ng BaPeq/g in soils, and from 0.02 to 0.16 ng BaPeq/m³ in air, indicative of high variability in this regional area. Human exposure to PAH varied between 976 and 42,877 ng BaPeq/d. When considering all exposure pathways, ILCR values were between 10⁻4 and 10⁻³. Considering only inhalation, ILCR values were between 10⁻6 and 10⁻5. The main risk seemed to arise from soil (either ingestion or inhalation of resuspended soil particles). The high spatial resolution of our environmental data allowed for detection of critical exposure levels at unexpected sites. Our results identified important areas where health studies on local populations need to be focused, and where environmental levels of PAH need to be monitored over time in order to protect human health.

A study protocol to evaluate the relationship between outdoor air pollution and pregnancy outcomes.

BACKGROUND: The present study protocol is designed to assess the relationship between outdoor air pollution and low birth weight and preterm births outcomes performing a semi-ecological analysis. Semi-ecological design studies are widely used to assess effects of air pollution in humans. In this type of analysis, health outcomes and covariates are measured in individuals and exposure assignments are usually based on air quality monitor stations. Therefore, estimating individual exposures are one of the major challenges when investigating these relationships with a semi-ecologic design. METHODS/DESIGN: Semi-ecologic study consisting of a retrospective cohort study with ecologic assignment of exposure is applied. Health outcomes and covariates are collected at Primary Health Care Center. Data from pregnant registry, clinical record and specific questionnaire administered orally to the mothers of children born in period 2007-2010 in Portuguese Alentejo Litoral region, are collected by the research team. Outdoor air pollution data are collected with a lichen diversity biomonitoring program, and individual pregnancy exposures are assessed with spatial geostatistical simulation, which provides the basis for uncertainty analysis of individual exposures. Awareness of outdoor air pollution uncertainty will improve validity of individual exposures assignments for further statistical analysis with multivariate regression models. DISCUSSION: Exposure misclassification is an issue of concern in semi-ecological design. In this study, personal exposures are assigned to each pregnant using geocoded addresses data. A stochastic simulation method is applied to lichen diversity values index measured at biomonitoring survey locations, in order to assess spatial uncertainty of lichen diversity value index at each geocoded address. These methods assume a model for spatial autocorrelation of exposure and provide a distribution of exposures in each study location. We believe that variability of simulated exposure values at geocoded addresses will improve knowledge on variability of exposures, improving therefore validity of individual exposures to input in posterior statistical analysis.

Spatial modeling of PAHs in lichens for fingerprinting of multisource atmospheric pollution.

PAHs are toxic compounds emitted by several anthropogenic sources, which have a great impact on human health. We show, for the first time, how spatial models based on PAHs intercepted by lichens can be used for fingerprinting multisource atmospheric pollution in a regional area. Urban-industrial areas showed the highest atmospheric deposition of PAHs followed by urban > industrial > agricultural > forest Multivariate analysis of lichen data showed, for the first time, a clear distinction between various sources of PAHs in the same area: urban are dominated by 4-ring PAHs, forest by 3-ring PAHs, and industrial by 5- and 6-ring PAHs or by 2-ring PAHs (petrogenic or pyrogenic, respectively). Heavy metals were also used for supporting the fingerprinting of PAH sources, reinforcing the industrial origin of 5- and 6-ring PAHs and revealing their particular nature. The spatial structure of the models for different PAHs seems to be dependent on the following factors: size and hydrophilic character of different PAHs, type of emission sources (point or nonpoint), and dispersion associated with particulates of different sizes. Based on the long-term integration of PAHs in lichens, these spatial models will significantly improve our knowledge on the impact of PAH chronic-exposure to humans and ecosystems.

The contribution of environmental biomonitoring with lichens to assess human exposure to dioxins.

The contribution of environmental biomonitoring with lichens to assess human exposure to dioxins was the main purpose of this work. For that, polychlorinated dibenzofurans (PCDD/F) were measured in 66 lichen sampling points. The obtained information significantly improved the basic knowledge on the environmental exposure to dioxins through distinction between effective control areas from areas with moderate atmospheric deposition. It allowed the integration of PCDD/F atmospheric deposition for much longer periods, allowing to relate low levels with long-term chronic effects on health. Thus, the production of high-resolution data on environmental exposure essential to perform reliable environmental health studies was possible. It was argued that PCDD/F in lichens may be used as spatial estimators of the potential risk of inhalation by the population present in the area. An example of the application of this data to select control and exposed areas for environmental health studies was presented.