Search

An increasing number of studies on non-contact vital sign detection using radar are now beginning to turn to data-driven neural network approaches rather than traditional signal-processing methods. However, there are few radar datasets available for deep learning due to the difficulty of acquiring and labeling the data, which require specialized equipment and physician collaboration. This paper presents a new model of heartbeat-induced chest wall motion (CWM) with the goal of generating a large amount of simulation data to support deep learning methods. An in-depth analysis of published CWM data collected by the VICON Infrared (IR) motion capture system and continuous wave (CW) radar system during respiratory hold was used to summarize the motion characteristics of each stage within a cardiac cycle. In combination with the physiological properties of the heartbeat, appropriate mathematical functions were selected to describe these movement properties. The model produced simulation data that closely matched the measured data as evaluated by dynamic time warping (DTW) and the root-mean-squared error (RMSE). By adjusting the model parameters, the heartbeat signals of different individuals were simulated. This will accelerate the application of data-driven deep learning methods in radar-based non-contact vital sign detection research and further advance the field. [ABSTRACT FROM AUTHOR]

In the pattern recognition domain, deep architectures are currently widely used and they have achieved fine results. However, these deep architectures make particular demands, especially in terms of their requirement for big datasets and GPU. Aiming to gain better results without deep networks, we propose a simplified algorithm framework using fusion features extracted from the salient areas of faces. Furthermore, the proposed algorithm has achieved a better result than some deep architectures. For extracting more effective features, this paper firstly defines the salient areas on the faces. This paper normalizes the salient areas of the same location in the faces to the same size; therefore, it can extracts more similar features from different subjects. LBP and HOG features are extracted from the salient areas, fusion features' dimensions are reduced by Principal Component Analysis (PCA) and we apply several classifiers to classify the six basic expressions at once. This paper proposes a salient areas definitude method which uses peak expressions frames compared with neutral faces. This paper also proposes and applies the idea of normalizing the salient areas to align the specific areas which express the different expressions. As a result, the salient areas found from different subjects are the same size. In addition, the gamma correction method is firstly applied on LBP features in our algorithm framework which improves our recognition rates significantly. By applying this algorithm framework, our research has gained state-of-the-art performances on CK+ database and JAFFE database. [ABSTRACT FROM AUTHOR]

Cloud computing is a rapidly expanding field. It allows users to access computer system resources as needed, particularly data storage and computational power, without managing them directly. This paper aims to create a data security model based on cryptography and steganography for data in cloud computing that seeks to reduce existing security and privacy concerns, such as data loss, data manipulation, and data theft. To identify the problem and determine its core cause, we studied various literature on existing cloud computing security models. This study utilizes design science research methodology. The design science research approach includes problem identification, requirements elicitation, artifact design and development, demonstration, and assessment. Design thinking and the Python programming language are used to build the artifact, and discussion about its working is represented using histograms, tables, and algorithms. This paper's output is a four-step data security model based on Rivest–Shamir–Adleman, Advanced Encryption Standard, and identity-based encryption algorithms alongside Least Significant Bit steganography. The four steps are data protection and security through encryption algorithms, steganography, data backup and recovery, and data sharing. This proposed approach ensures more cloud data redundancy, flexibility, efficiency, and security by protecting data confidentiality, privacy, and integrity from attackers. [ABSTRACT FROM AUTHOR]

Optical fiber communications are essential for all types of long- and short-distance transmissions. The aim of this paper is to analyze the previously presented security risks and, based on measurements, provide the risk level evaluation. The major risk is the possibility of inserting a splitter into the optical distribution network and capturing a portion of the entire spectrum, i.e., all channels in the optical fiber. Another significant security risk is crosstalk on multiplexers in networks with wavelength division multiplexing. The paper covers the macrobend attenuation evaluation of fiber and back-reflection measurements. Based on the measurements, risks were evaluated for both point-to-point and point-to-multipoint networks and, lastly, the paper covers crosstalk measurements of an optomechanical switch. Finally, all individual risks are evaluated according to the severity, and a proposal for risk minimization is provided. [ABSTRACT FROM AUTHOR]

This paper presents a novel CNN-based architecture, referred to as Q-Net, to learn local feature descriptors that are useful for matching image patches from two different spectral bands. Given correctly matched and non-matching cross-spectral image pairs, a quadruplet network is trained to map input image patches to a common Euclidean space, regardless of the input spectral band. Our approach is inspired by the recent success of triplet networks in the visible spectrum, but adapted for cross-spectral scenarios, where, for each matching pair, there are always two possible non-matching patches: one for each spectrum. Experimental evaluations on a public cross-spectral VIS-NIR dataset shows that the proposed approach improves the state-of-the-art. Moreover, the proposed technique can also be used in mono-spectral settings, obtaining a similar performance to triplet network descriptors, but requiring less training data. [ABSTRACT FROM AUTHOR]

In this paper, compressed sensing (CS) theory is utilized in a medium-access control (MAC) scheme for wireless sensor networks (WSNs). We propose a new, cross-layer compressed sensing medium-access control (CL CS-MAC) scheme, combining the physical layer and data link layer, where the wireless transmission in physical layer is considered as a compress process of requested packets in a data link layer according to compressed sensing (CS) theory. We first introduced using compressive complex requests to identify the exact active sensor nodes, which makes the scheme more efficient. Moreover, because the reconstruction process is executed in a complex field of a physical layer, where no bit and frame synchronizations are needed, the asynchronous and random requests scheme can be implemented without synchronization payload. We set up a testbed based on software-defined radio (SDR) to implement the proposed CL CS-MAC scheme practically and to demonstrate the validation. For large-scale WSNs, the simulation results show that the proposed CL CS-MAC scheme provides higher throughput and robustness than the carrier sense multiple access (CSMA) and compressed sensing medium-access control (CS-MAC) schemes. [ABSTRACT FROM AUTHOR]

This paper presents the Monitoring and Discovery Framework of the Self-Organized Network Management in Virtualized and Software Defined Networks SELFNET project. This design takes into account the scalability and flexibility requirements needed by 5G infrastructures. In this context, the present framework focuses on gathering and storing the information (low-level metrics) related to physical and virtual devices, cloud environments, flow metrics, SDN traffic and sensors. Similarly, it provides the monitoring data as a generic information source in order to allow the correlation and aggregation tasks. Our design enables the collection and storing of information provided by all the underlying SELFNET sublayers, including the dynamically onboarded and instantiated SDN/NFV Apps, also known as SELFNET sensors. [ABSTRACT FROM AUTHOR]

The operation and maintenance of buildings has seen several advances in recent years. Multiple information and communication technology (ICT) solutions have been introduced to better manage building maintenance. However, maintenance practices in buildings remain less efficient and lead to significant energy waste. In this paper, a predictive maintenance framework based on machine learning techniques is proposed. This framework aims to provide guidelines to implement predictive maintenance for building installations. The framework is organised into five steps: data collection, data processing, model development, fault notification and model improvement. A sport facility was selected as a case study in this work to demonstrate the framework. Data were collected from different heating ventilation and air conditioning (HVAC) installations using Internet of Things (IoT) devices and a building automation system (BAS). Then, a deep learning model was used to predict failures. The case study showed the potential of this framework to predict failures. However, multiple obstacles and barriers were observed related to data availability and feedback collection. The overall results of this paper can help to provide guidelines for scientists and practitioners to implement predictive maintenance approaches in buildings. [ABSTRACT FROM AUTHOR]

Real-time queue length information is an important input for many traffic applications. This paper presents a novel method for real-time queue length detection with roadside LiDAR data. Vehicles on the road were continuously tracked with the LiDAR data processing procedures (including background filtering, point clustering, object classification, lane identification and object association). A detailed method to identify the vehicle at the end of the queue considering the occlusion issue and package loss issue was documented in this study. The proposed method can provide real-time queue length information. The performance of the proposed queue length detection method was evaluated with the ground-truth data collected from three sites in Reno, Nevada. Results show the proposed method can achieve an average of 98% accuracy at the six investigated sites. The errors in the queue length detection were also diagnosed. [ABSTRACT FROM AUTHOR]

Deep Learning (DL), a successful promising approach for discriminative and generative tasks, has recently proved its high potential in 2D medical imaging analysis; however, physiological data in the form of 1D signals have yet to be beneficially exploited from this novel approach to fulfil the desired medical tasks. Therefore, in this paper we survey the latest scientific research on deep learning in physiological signal data such as electromyogram (EMG), electrocardiogram (ECG), electroencephalogram (EEG), and electrooculogram (EOG). We found 147 papers published between January 2018 and October 2019 inclusive from various journals and publishers. The objective of this paper is to conduct a detailed study to comprehend, categorize, and compare the key parameters of the deep-learning approaches that have been used in physiological signal analysis for various medical applications. The key parameters of deep-learning approach that we review are the input data type, deep-learning task, deep-learning model, training architecture, and dataset sources. Those are the main key parameters that affect system performance. We taxonomize the research works using deep-learning method in physiological signal analysis based on: (1) physiological signal data perspective, such as data modality and medical application; and (2) deep-learning concept perspective such as training architecture and dataset sources. [ABSTRACT FROM AUTHOR]

The '15-minute city' concept is emerging as a potent urban regeneration model in post-pandemic cities, offering new vantage points on liveability and urban health. While the concept is primarily geared towards rethinking urban morphologies, it can be furthered via the adoption of Smart Cities network technologies to provide tailored pathways to respond to contextualised challenges through the advent of data mining and processing to better inform urban decision-making processes. We argue that the '15-minute city' concept can value-add from Smart City network technologies in particular through Digital Twins, Internet of Things (IoT), and 6G. The data gathered by these technologies, and processed via Machine Learning techniques, can unveil new patterns to understand the characteristics of urban fabrics. Collectively, those dimensions, unpacked to support the '15-minute city' concept, can provide new opportunities to redefine agendas to better respond to economic and societal needs as well as align more closely with environmental commitments, including the United Nations' Sustainable Development Goal 11 and the New Urban Agenda. This perspective paper presents new sets of opportunities for cities arguing that these new connectivities should be explored now so that appropriate protocols can be devised and so that urban agendas can be recalibrated to prepare for upcoming technology advances, opening new pathways for urban regeneration and resilience crafting. [ABSTRACT FROM AUTHOR]

This paper presents an intelligent system allowing handicapped aphasiacs to perform basic communication tasks. It has the following three key features: (1) A 6-sensor data glove measures the finger gestures of a patient in terms of the bending degrees of his fingers. (2) A finger language recognition subsystem recognizes language components from the finger gestures. It employs multiple regression analysis to automatically extract proper finger features so that the recognition model can be fast and correctly constructed by a radial basis function neural network. (3) A coordinate-indexed virtual keyboard allows the users to directly access the letters on the keyboard at a practical speed. The system serves as a viable tool for natural and affordable communication for handicapped aphasiacs through continuous finger language input. [ABSTRACT FROM AUTHOR]

Flash memory has become a more widespread storage medium for modern wireless devices because of its effective characteristics like non-volatility, small size, light weight, fast access speed, shock resistance, high reliability and low power consumption. Sensor nodes are highly resource constrained in terms of limited processing speed, runtime memory, persistent storage, communication bandwidth and finite energy. Therefore, for wireless sensor networks supporting sense, store, merge and send schemes, an efficient and reliable file system is highly required with consideration of sensor node constraints. In this paper, we propose a novel log structured external NAND flash memory based file system, called Proceeding to Intelligent service oriented memorY Allocation for flash based data centric Sensor devices in wireless sensor networks (PIYAS). This is the extended version of our previously proposed PIYA [1]. The main goals of the PIYAS scheme are to achieve instant mounting and reduced SRAM space by keeping memory mapping information to a very low size of and to provide high query response throughput by allocation of memory to the sensor data by network business rules. The scheme intelligently samples and stores the raw data and provides high in-network data availability by keeping the aggregate data for a longer period of time than any other scheme has done before. We propose effective garbage collection and wear-leveling schemes as well. The experimental results show that PIYAS is an optimized memory management scheme allowing high performance for wireless sensor networks. [ABSTRACT FROM AUTHOR]

Wi-Fi uploading is considered an effective method for offloading the traffic of cellular networks generated by the data uploading process of mobile crowd sensing applications. However, previously proposed Wi-Fi uploading schemes mainly focus on optimizing one performance objective: the offloaded cellular traffic or the reduced uploading cost. In this paper, we propose an Intelligent Data Uploading Selection Mechanism (IDUSM) to realize a trade-off between the offloaded traffic of cellular networks and participants' uploading cost considering the differences among participants' data plans and direct and indirect opportunistic transmissions. The mechanism first helps the source participant choose an appropriate data uploading manner based on the proposed probability prediction model, and then optimizes its performance objective for the chosen data uploading manner. In IDUSM, our proposed probability prediction model precisely predicts a participant's mobility from spatial and temporal aspects, and we decrease data redundancy produced in the Wi-Fi offloading process to reduce waste of participants' limited resources (e.g., storage, battery). Simulation results show that the offloading efficiency of our proposed IDUSM is (56.54 × 10 − 7) , and the value is the highest among the other three Wi-Fi offloading mechanisms. Meanwhile, the offloading ratio and uploading cost of IDUSM are respectively 52.1% and (6.79 × 10 3) . Compared with other three Wi-Fi offloading mechanisms, it realized a trade-off between the offloading ratio and the uploading cost. [ABSTRACT FROM AUTHOR]

This paper looks at the development and implementation of the additional secondary phase factor (ASF) real-time correction system for the Automatic Identification System (AIS) signal. A large number of test data were collected using the developed ASF correction system and the propagation characteristics of the AIS signal that transmits at sea and the ASF real-time correction algorithm of the AIS signal were analyzed and verified. Accounting for the different hardware of the receivers in the land-based positioning system and the variation of the actual environmental factors, the ASF correction system corrects original measurements of positioning receivers in real time and provides corrected positioning accuracy within 10 m. [ABSTRACT FROM AUTHOR]

Patients discharged from hospitals after an inpatient course of medical treatment for any ailment or traumatic injury that results in disabling conditions and are rendered mobility impaired require ongoing systematic sports and exercise programs to maintain healthy lifestyles. Under such circumstances, a rehabilitation exercise and sports center, accessible throughout local communities, is critical for promoting beneficial living and community participation for these individuals with disabilities. An innovative data-driven system equipped with state-of-the-art smart and digital equipment, set up in architecturally barrier-free infrastructures, is essential for these individuals to promote health maintenance and overcome secondary medical complications following an acute inpatient hospitalization or suboptimal rehabilitation. A federally funded collaborative research and development (R&D) program proposes to build a multi-ministerial data-driven system of exercise programs using a smart digital living lab as a platform to provide pilot services in physical education and counseling with exercise and sports programs for this patient population. We describe the social and critical aspects of rehabilitating such a population of patients by presenting a full study protocol. A modified sub-dataset of the previously generated 280-item full dataset is applied using a data-collecting system—"The Elephant"—as an example of how data acquisition will be achieved to assess the effects of lifestyle rehabilitative exercise programs for people with disabilities. [ABSTRACT FROM AUTHOR]

This paper presents a conceptual framework for the optimization of environmental sustainability in engineering projects, both for products and industrial facilities or processes. The main objective of this work is to propose a conceptual framework to help researchers to approach optimization under the criteria of sustainability of engineering projects, making use of current Machine Learning techniques. For the development of this conceptual framework, a bibliographic search has been carried out on the Web of Science. From the selected documents and through a hermeneutic procedure the texts have been analyzed and the conceptual framework has been carried out. A graphic representation pyramid shape is shown to clearly define the variables of the proposed conceptual framework and their relationships. The conceptual framework consists of 5 dimensions; its acronym is ADAPTS. In the base are: (1) the Application to which it is intended, (2) the available DAta, (3) the APproach under which it is operated, and (4) the machine learning Tool used. At the top of the pyramid, (5) the necessary Sensing. A study case is proposed to show its applicability. This work is part of a broader line of research, in terms of optimization under sustainability criteria. [ABSTRACT FROM AUTHOR]

Activity recognition, a key component in pervasive healthcare monitoring, relies on classification algorithms that require labeled data of individuals performing the activity of interest to train accurate models. Labeling data can be performed in a lab setting where an individual enacts the activity under controlled conditions. The ubiquity of mobile and wearable sensors allows the collection of large datasets from individuals performing activities in naturalistic conditions. Gathering accurate data labels for activity recognition is typically an expensive and time-consuming process. In this paper we present two novel approaches for semi-automated online data labeling performed by the individual executing the activity of interest. The approaches have been designed to address two of the limitations of self-annotation: (i) The burden on the user performing and annotating the activity, and (ii) the lack of accuracy due to the user labeling the data minutes or hours after the completion of an activity. The first approach is based on the recognition of subtle finger gestures performed in response to a data-labeling query. The second approach focuses on labeling activities that have an auditory manifestation and uses a classifier to have an initial estimation of the activity, and a conversational agent to ask the participant for clarification or for additional data. Both approaches are described, evaluated in controlled experiments to assess their feasibility and their advantages and limitations are discussed. Results show that while both studies have limitations, they achieve 80% to 90% precision. [ABSTRACT FROM AUTHOR]

This review paper discusses the trends and projections for wearable technology in the consumer sports sector (excluding professional sport). Analyzing the role of wearable technology for different users and why there is such a need for these devices in everyday lives. It shows how different sensors are influential in delivering a variety of readings that are useful in many ways regarding sport attributes. Wearables are increasing in function, and through integrating technology, users are gathering more data about themselves. The amount of wearable technology available is broad, each having its own role to play in different industries. Inertial measuring unit (IMU) and Global Positioning System (GPS) sensors are predominantly present in sport wearables but can be programmed for different needs. In this review, the differences are displayed to show which sensors are compatible and which ones can evolve sensor technology for sport applications. [ABSTRACT FROM AUTHOR]

Sensor drift is the most challenging problem in gas sensing at present. We propose a novel two-dimensional classifier ensemble strategy to solve the gas discrimination problem, regardless of the gas concentration, with high accuracy over extended periods of time. This strategy is appropriate for multi-class classifiers that consist of combinations of pairwise classifiers, such as support vector machines. We compare the performance of the strategy with those of competing methods in an experiment based on a public dataset that was compiled over a period of three years. The experimental results demonstrate that the two-dimensional ensemble outperforms the other methods considered. Furthermore, we propose a pre-aging process inspired by that applied to the sensors to improve the stability of the classifier ensemble. The experimental results demonstrate that the weight of each multi-class classifier model in the ensemble remains fairly static before and after the addition of new classifier models to the ensemble, when a pre-aging procedure is applied. [ABSTRACT FROM AUTHOR]

Tracking the locations and identities of moving targets in the surveillance area of wireless sensor networks is studied. In order to not rely on high-cost sensors that have been used in previous researches, we propose the integrated localization and classification based on semi-supervised learning, which uses both labeled and unlabeled data obtained from low-cost distributed sensor network. In our setting, labeled data are obtained by seismic and PIR sensors that contain information about the types of the targets. Unlabeled data are generated from the RF signal strength by applying Gaussian process, which represents the probability of predicted target locations. Finally, by using classified unlabeled data produced by semi-supervised learning, identities and locations of multiple targets are estimated. In addition, we consider a case when the labeled data are absent, which can happen due to fault or lack of the deployed sensor nodes and communication failure. We overcome this situation by defining artificial labeled data utilizing characteristics of support vector machine, which provides information on the importance of each training data point. Experimental results demonstrate the accuracy of the proposed tracking algorithm and its robustness to the absence of the labeled data thanks to the artificial labeled data. [ABSTRACT FROM AUTHOR]

Global solar broadband irradiance on a planar surface is measured at weather stations by pyranometers. In the case of the present research, solar radiation values from nine meteorological stations of the MeteoGalicia real-time observational network, captured and stored every ten minutes, are considered. In this kind of record, the lack of data and/or the presence of wrong values adversely affects any time series study. Consequently, when this occurs, a data imputation process must be performed in order to replace missing data with estimated values. This paper aims to evaluate the multivariate imputation of ten-minute scale data by means of the chained equations method (MICE). This method allows the network itself to impute the missing or wrong data of a solar radiation sensor, by using either all or just a group of the measurements of the remaining sensors. Very good results have been obtained with the MICE method in comparison with other methods employed in this field such as Inverse Distance Weighting (IDW) and Multiple Linear Regression (MLR). The average RMSE value of the predictions for the MICE algorithm was 13.37% while that for the MLR it was 28.19%, and 31.68% for the IDW. [ABSTRACT FROM AUTHOR]

Visualization of sensor networks, data, and metadata is becoming one of the most pivotal aspects of the structural health monitoring (SHM) process. Without the ability to communicate efficiently and effectively between disparate groups working on a project, an SHM system can be underused, misunderstood, or even abandoned. For this reason, this work seeks to evaluate visualization techniques in the field, identify flaws in current practices, and devise a new method for visualizing and accessing SHM data and metadata in 3D. More precisely, the work presented here reflects a method and digital workflow for integrating SHM sensor networks, data, and metadata into a virtual reality environment by combining spherical imaging and informational modeling. Both intuitive and interactive, this method fosters communication on a project enabling diverse practitioners of SHM to efficiently consult and use the sensor networks, data, and metadata. The method is presented through its implementation on a case study, Streicker Bridge at Princeton University campus. To illustrate the efficiency of the new method, the time and data file size were compared to other potential methods used for visualizing and accessing SHM sensor networks, data, and metadata in 3D. Additionally, feedback from civil engineering students familiar with SHM is used for validation. Recommendations on how different groups working together on an SHM project can create SHM virtual environment and convey data to proper audiences, are also included. [ABSTRACT FROM AUTHOR]

As the rapid growth of mobile users and Internet-of-Everything devices will continue in the upcoming decade, more and more network capacity will be needed to accommodate such a constant increase in data volumes (DVs). To satisfy such a vast DV increase, the implementation of the fifth-generation (5G) and future sixth-generation (6G) mobile networks will be based on heterogeneous networks (HetNets) composed of macro base stations (BSs) dedicated to ensuring basic signal coverage and capacity, and small BSs dedicated to satisfying capacity for increased DVs at locations of traffic hotspots. An approach that can accommodate constantly increasing DVs is based on adding additional capacity in the network through the deployment of new BSs as DV increases. Such an approach represents an implementation challenge to mobile network operators (MNOs), which is reflected in the increased power consumption of the radio access part of the mobile network and degradation of network energy efficiency (EE). In this study, the impact of the expected increase of DVs through the 2020s on the EE of the 5G radio access network (RAN) was analyzed by using standardized data and coverage EE metrics. An analysis was performed for five different macro and small 5G BS implementation and operation scenarios and for rural, urban, dense-urban and indoor-hotspot device density classes (areas). The results of analyses reveal a strong influence of increasing DV trends on standardized data and coverage EE metrics of 5G HetNets. For every device density class characterized with increased DVs, we here elaborate on the process of achieving the best and worse combination of data and coverage EE metrics for each of the analyzed 5G BSs deployment and operation approaches. This elaboration is further extended on the analyses of the impact of 5G RAN instant power consumption and 5G RAN yearly energy consumption on values of standardized EE metrics. The presented analyses can serve as a reference in the selection of the most appropriate 5G BS deployment and operation approach, which will simultaneously ensure the transfer of permanently increasing DVs in a specific device density class and the highest possible levels of data and coverage EE metrics. [ABSTRACT FROM AUTHOR]

Infiltration-route analysis is a military application of geospatial information system (GIS) technology. In order to find susceptible routes, optimal-path-searching algorithms are applied to minimize the cost function, which is the summed result of detection probability. The cost function was determined according to the thermal observation device (TOD) detection probability, the viewshed analysis results, and two feature layers extracted from the vector product interim terrain data. The detection probability is computed and recorded for an individual cell (50 m × 50 m), and the optimal infiltration routes are determined with A* algorithm by minimizing the summed costs on the routes from a start point to an end point. In the present study, in order to simulate the dynamic nature of a realworld problem, one thousand cost surfaces in the GIS environment were generated with randomly located TODs and randomly selected infiltration start points. Accordingly, one thousand sets of vulnerable routes for infiltration purposes could be found, which could be accumulated and presented as an infiltration vulnerability map. This application can be further utilized for both optimal infiltration routing and surveillance network design. Indeed, dynamic simulation in the GIS environment is considered to be a powerful and practical solution for optimization problems. A similar approach can be applied to the dynamic optimal routing for civil infrastructure, which requires consideration of terrain-related constraints and cost functions. [ABSTRACT FROM AUTHOR]

Driver inattention is a major contributor to road crashes. The emerging of new driver monitoring systems represents an opportunity for researchers to explore new data sources to understand driver inattention, even if the technology was not developed with this purpose in mind. This study is based on retrospective data obtained from two driver monitoring systems to study distraction and drowsiness risk factors. The data includes information about the trips performed by 330 drivers and corresponding distraction and drowsiness alerts emitted by the systems. The drivers' historical travel data allowed defining two groups with different mobility patterns (short-distance and long-distance drivers) through a cluster analysis. Then, the impacts of the driver's profile and trip characteristics (e.g., driving time, average speed, and breaking time and frequency) on inattention were analyzed using ordered probit models. The results show that long-distance drivers, typically associated with professionals, are less prone to distraction and drowsiness than short-distance drivers. The driving time increases the probability of inattention, while the breaking frequency is more important to mitigate inattention than the breaking time. Higher average speeds increase the inattention risk, being associated with road facilities featuring a monotonous driving environment. [ABSTRACT FROM AUTHOR]

For autonomous vehicles (AV), the ability to share information about their surroundings is crucial. With Level 4 and 5 autonomy in sight, solving the challenge of organization and efficient storing of data, coming from these connected platforms, becomes paramount. Research done up to now has been mostly focused on communication and network layers of V2X (Vehicle-to-Everything) data sharing. However, there is a gap when it comes to the data layer. Limited attention has been paid to the ontology development in the automotive domain. More specifically, the way to integrate sensor data and geospatial data efficiently is missing. Therefore, we proposed to develop a new Connected Traffic Data Ontology (CTDO) on the foundations of Sensor, Observation, Sample, and Actuator (SOSA) ontology, to provide a more suitable ontology for large volumes of time-sensitive data coming from multi-sensory platforms, like connected vehicles, as the first step in closing the existing research gap. Additionally, as this research aims to further extend the CTDO in the future, a possible way to map to the CTDO with ontologies that represent road infrastructure has been presented. Finally, new CTDO ontology was benchmarked against SOSA, and better memory performance and query execution speeds have been confirmed. [ABSTRACT FROM AUTHOR]

Accounting for realistic observation errors is a known bottleneck in data assimilation, because dealing with error correlations is complex. Following a previous study on this subject, we propose to use multiscale modelling, more precisely wavelet transform, to address this question. This study aims to investigate the problem further by addressing two issues arising in real-life data assimilation: how to deal with partially missing data (e.g., concealed by an obstacle between the sensor and the observed system), and how to solve convergence issues associated with complex observation error covariance matrices? Two adjustments relying on wavelets modelling are proposed to deal with those, and offer significant improvements. The first one consists of adjusting the variance coefficients in the frequency domain to account for masked information. The second one consists of a gradual assimilation of frequencies. Both of these fully rely on the multiscale properties associated with wavelet covariance modelling. Numerical results on twin experiments show that multiscale modelling is a promising tool to account for correlations in observation errors in realistic applications. [ABSTRACT FROM AUTHOR]

Few methods for classifying physical activity from accelerometer data have been tested using an independent dataset for cross-validation, and even fewer using multiple independent datasets. The aim of this study was to evaluate whether unsupervised machine learning was a viable approach for the development of a reusable clustering model that was generalisable to independent datasets. We used two labelled adult laboratory datasets to generate a k-means clustering model. To assess its generalised application, we applied the stored clustering model to three independent labelled datasets: two laboratory and one free-living. Based on the development labelled data, the ten clusters were collapsed into four activity categories: sedentary, standing/mixed/slow ambulatory, brisk ambulatory, and running. The percentages of each activity type contained in these categories were 89%, 83%, 78%, and 96%, respectively. In the laboratory independent datasets, the consistency of activity types within the clusters dropped, but remained above 70% for the sedentary clusters, and 85% for the running and ambulatory clusters. Acceleration features were similar within each cluster across samples. The clusters created reflected activity types known to be associated with health and were reasonably robust when applied to diverse independent datasets. This suggests that an unsupervised approach is potentially useful for analysing free-living accelerometer data. [ABSTRACT FROM AUTHOR]

With the growing number of mobile devices receiving daily notifications, it is necessary to manage the variety of information produced. New smart devices are developed every day with the ability to generate, send, and display messages about their status, data, and information about other devices. Consequently, the number of notifications received by a user is increasing and their tolerance may decrease in a short time. With this, it is necessary to develop a management system and notification controls. In this context, this work proposes a notification and alert management system called PRISER. Its focus is on user profiles and environments, applying data privacy criteria. [ABSTRACT FROM AUTHOR]