A TinyML Soft-Sensor Approach for Low-Cost Detection and Monitoring of Vehicular Emissions.

Vehicles are the major source of air pollution in modern cities, emitting excessive levels of CO2 and other noxious gases. Exploiting the OBD-II interface available on most vehicles, the continuous emission of such pollutants can be indirectly measured over time, although accuracy has been an important design issue when performing this task due the nature of the retrieved data. In this scenario, soft-sensor approaches can be adopted to process engine combustion data such as fuel injection and mass air flow, processing them to estimate pollution and transmitting the results for further analyses. Therefore, this article proposes a soft-sensor solution based on an embedded system designed to retrieve data from vehicles through their OBD-II interface, processing different inputs to provide estimated values of CO2 emissions over time. According to the type of data provided by the vehicle, two different algorithms are defined, and each follows a comprehensive mathematical formulation. Moreover, an unsupervised TinyML approach is also derived to remove outliers data when processing the computed data stream, improving the accuracy of the soft sensor as a whole while not requiring any interaction with cloud-based servers to operate. Initial results for an embedded implementation on the Freematics ONE+ board have shown the proposal's feasibility with an acquisition frequency equal to 1Hz and emission granularity measure of gCO2/km.

Discovering temporal scientometric knowledge in COVID-19 scholarly production.

The mapping and analysis of scientific knowledge makes it possible to identify the dynamics and/or growth of a particular field of research or to support strategic decisions related to different research entities, based on bibliometric and/or scientometric indicators. However, with the exponential growth of scientific production, a systematic and data-oriented approach to the analysis of this large set of productions becomes increasingly essential. Thus, in this work, a data-oriented methodology was proposed, combining Data Analysis, Machine Learning and Complex Network Analysis techniques, and Data Version Control (DVC) tool, for the extraction of implicit knowledge in scientific production bases. In addition, the approach was validated through a case study in a COVID-19 manuscripts dataset, which had 199,895 articles published on arXiv, bioRxiv, medRxiv, PubMed and Scopus databases. The results suggest the feasibility of the proposed methodology, indicating the most active countries and the most explored themes in each period of the pandemic. Therefore, this study has the potential to instrument and expand strategic decisions by the scientific community, aiming at extracting knowledge that supports the fight against the COVID-19 pandemic.

An Evolving TinyML Compression Algorithm for IoT Environments Based on Data Eccentricity.

Currently, the applications of the Internet of Things (IoT) generate a large amount of sensor data at a very high pace, making it a challenge to collect and store the data. This scenario brings about the need for effective data compression algorithms to make the data manageable among tiny and battery-powered devices and, more importantly, shareable across the network. Additionally, considering that, very often, wireless communications (e.g., low-power wide-area networks) are adopted to connect field devices, user payload compression can also provide benefits derived from better spectrum usage, which in turn can result in advantages for high-density application scenarios. As a result of this increase in the number of connected devices, a new concept has emerged, called TinyML. It enables the use of machine learning on tiny, computationally restrained devices. This allows intelligent devices to analyze and interpret data locally and in real time. Therefore, this work presents a new data compression solution (algorithm) for the IoT that leverages the TinyML perspective. The new approach is called the Tiny Anomaly Compressor (TAC) and is based on data eccentricity. TAC does not require previously established mathematical models or any assumptions about the underlying data distribution. In order to test the effectiveness of the proposed solution and validate it, a comparative analysis was performed on two real-world datasets with two other algorithms from the literature (namely Swing Door Trending (SDT) and the Discrete Cosine Transform (DCT)). It was found that the TAC algorithm showed promising results, achieving a maximum compression rate of 98.33%. Additionally, it also surpassed the two other models regarding the compression error and peak signal-to-noise ratio in all cases.

On the development of flexible mobile multi-sensor units based on open-source hardware platforms and a reference framework.

For some IoT applications, mobile entities are considered as the main source of sensed data, requiring the attachment of sensor modules on them. The endowing of sensing capabilities to such mobile entities can be performed in different ways, but the adoption of a reference hardware framework can bring a series of advantages, specially in dynamic complex scenarios. This article exploits the MSensorMob2 multi-sensor hardware framework for monitoring in areas with disconnection periods, comprising sensing, transmission and reconfiguration functions. Comprehensive analyses on multiple open-source hardware platforms are conducted, assessing their costs, deployment constraints and performance issues when implementing this development framework.

Artificial Mercosur license plates dataset.

Mercosur (a.k.a. Mercosul) is a trade bloc comprising five South American countries. In 2018, a unified Mercosur license plate model was rolled out. Access to large volumes of ground truth Mercosur license plates with sufficient presentation variety is a significant challenge for training supervised models for license plate detection (LPD) in automatic license plate recognition (ALPR) systems. To address this problem, a Mercosur license plate generator was developed to generate artificial license plate images meeting the new standard with sufficient variety for ALPR training purposes. This includes images with variation due to occlusions and environmental conditions. An embedded system was developed for detecting legacy license plates in images of real scenarios and overwriting these with artificially generated Mercosur license plates. This data set comprises 3,829 images of vehicles with synthetic license plates that meet the new Mercosur standard in real scenarios, and equivalent number of text files containing label information for the images, all organized in a CSV file with compiled image file paths and associated labels.

COVID-19: A scholarly production dataset report for research analysis.

COVID-2019 has been recognized as a global threat, and several studies are being conducted in order to contribute to the fight and prevention of this pandemic. This work presents a scholarly production dataset focused on COVID-19, providing an overview of scientific research activities, making it possible to identify countries, scientists and research groups most active in this task force to combat the coronavirus disease. The dataset is composed of 40,212 records of articles' metadata collected from Scopus, PubMed, arXiv and bioRxiv databases from January 2019 to July 2020. Those data were extracted by using the techniques of Python Web Scraping and preprocessed with Pandas Data Wrangling. In addition, the pipeline to preprocess and generate the dataset are versioned with the Data Version Control tool (DVC) and are thus easily reproducible and auditable.

#StayHome: Monitoring and benchmarking social isolation trends in Caruaru and the Região Metropolitana do Recife during the COVID-19 pandemic.

This technical report presents information related to the Social Isolation Index (SII) of the city of Caruaru, Pernambuco, Brazil. The data was provided by In Loco, a technology startup that has collected the movement of around 60 million Brazilians through cell phone location.

Dataset for country profile and mobility analysis in the assessment of COVID-19 pandemic.

Understanding the COVID-19 pandemic is a multidisciplinary effort that requires a significant number of variables. This dataset comprises (i) sociodemographic characteristics, compiled from 35 datasets obtained at UN Data; (ii) mobility metrics that can assist the analysis of social distancing, from Google Community Mobility Reports and; (iii) daily counts of cases and deaths by COVID-19, from the European Centre for Disease Prevention and Control and the Johns Hopkins University Center for Systems Science and Engineering. This unified dataset ranges from February 15, 2020 to May 7, 2020, a total of 83 days, and is provided as a collection of time series for 131 countries with 192 variables. The pipeline to preprocess and generate the dataset, along with the dataset itself, are versioned with the Data Version Control tool (DVC) and are thus easily reproducible.

Data set for automatic detection of online misogynistic speech.

The data set is composed of 2285 definitions posted on the Urban Dictionary platform from 1999 to May 2016. The data was classified as misogynistic and non-misogynistic by three independent researchers with domain knowledge. The data set is available in public repository in a table containing two columns: the text-based definition from Urban Dictionary and its respective classification (1 for misogynistic and 0 for non-misogynistic).

Extracting Value from Industrial Alarms and Events: A Data-Driven Approach Based on Exploratory Data Analysis.

Alarm and event logs are an immense but latent source of knowledge commonly undervalued in industry. Though, the current massive data-exchange, high efficiency and strong competitiveness landscape, boosted by Industry 4.0 and IIoT (Industrial Internet of Things) paradigms, does not accommodate such a data misuse and demands more incisive approaches when analyzing industrial data. Advances in Data Science and Big Data (or more precisely, Industrial Big Data) have been enabling novel approaches in data analysis which can be great allies in extracting hitherto hidden information from plant operation data. Coping with that, this work proposes the use of Exploratory Data Analysis (EDA) as a promising data-driven approach to pave industrial alarm and event analysis. This approach proved to be fully able to increase industrial perception by extracting insights and valuable information from real-world industrial data without making prior assumptions.

Accelerometer-Based Human Fall Detection Using Convolutional Neural Networks.

Human falls are a global public health issue resulting in over 37.3 million severe injuries and 646,000 deaths yearly. Falls result in direct financial cost to health systems and indirectly to society productivity. Unsurprisingly, human fall detection and prevention are a major focus of health research. In this article, we consider deep learning for fall detection in an IoT and fog computing environment. We propose a Convolutional Neural Network composed of three convolutional layers, two maxpool, and three fully-connected layers as our deep learning model. We evaluate its performance using three open data sets and against extant research. Our approach for resolving dimensionality and modelling simplicity issues is outlined. Accuracy, precision, sensitivity, specificity, and the Matthews Correlation Coefficient are used to evaluate performance. The best results are achieved when using data augmentation during the training process. The paper concludes with a discussion of challenges and future directions for research in this domain.

An Effective Extension of Anti-Collision Protocol for RFID in the Industrial Internet of Things (IIoT).

The Industrial Internet of Things (IIoT) is often presented as a concept that is significantly changing industry, yet continuous improvements in the identification and automation of objects are still required. Such improvements are related to communication speed, security, and reliability, critical attributes for industrial environments. In this context, the radio-frequency identification (RFID) systems present some issues related to frame collision when there are several tags transmitting data. The dynamic framed-slotted ALOHA (DFSA) is a widely used algorithm to solve collision problems in RFID systems. DFSA dynamically adjusts the frame length based on estimations of the number of labels that have competed for slots in the previous frame. Thus, the accuracy of the estimator is directly related to the label identification performance. In the literature, there are several estimators proposed to improve labels identification accuracy. However, they are not efficient when considering a large tag population, requiring a considerable amount of computational resources to perform the identification. In this context, this work proposes an estimator, which can efficiently identify a large number of labels without requiring additional computational resources. Through a set of simulations, the results demonstrate that the proposed estimator has a nearly ideal channel usage efficiency of 36.1%, which is the maximum efficiency of the DFSA protocol.

A Customer Feedback Platform for Vehicle Manufacturing Compliant with Industry 4.0 Vision.

In the last decade, the growth of the automotive market with the aid of technologies has been notable for the economic, automotive and technological sectors. Alongside this growing recognition, the so called Internet of Intelligent Vehicles (IoIV) emerges as an evolution of the Internet of Things (IoT) applied to the automotive sector. Closely related to IoIV, emerges the concept of Industrial Internet of Things (IIoT), which is the current revolution seen in industrial automation. IIoT, in its turn, relates to the concept of Industry 4.0, that is used to represent the current Industrial Revolution. This revolution, however, involves different areas: from manufacturing to healthcare. The Industry 4.0 can create value during the entire product lifecycle, promoting customer feedback, that is, having information about the product history throughout it is life. In this way, the automatic communication between vehicle and factory was facilitated, allowing the accomplishment of different analysis regarding vehicles, such as the identification of a behavioral pattern through historical driver usage, fuel consumption, maintenance indicators, so on. Thus, allowing the prevention of critical issues and undesired behaviors, since the automakers lose contact with the vehicle after the purchase. Therefore, this paper aims to propose a customer feedback platform for vehicle manufacturing in Industry 4.0 context, capable of collecting and analyzing, through an OBD-II (On-Board Diagnostics) scanner, the sensors available by vehicles, with the purpose of assisting in the management, prevention, and mitigation of different vehicular problems. An intercontinental evaluation conducted between Brazil and Italy locations shown the feasibility of platform and the potential to use in order to improve the vehicle manufacturing process.

Routing and Scheduling Algorithms for WirelessHARTNetworks: A Survey.

Wireless communication is a trend nowadays for the industrial environment. A number of different technologies have emerged as solutions satisfying strict industrial requirements (e.g., WirelessHART, ISA100.11a, WIA-PA). As the industrial environment presents a vast range of applications, adopting an adequate solution for each case is vital to obtain good performance of the system. In this context, the routing and scheduling schemes associated with these technologies have a direct impact on important features, like latency and energy consumption. This situation has led to the development of a vast number of routing and scheduling schemes. In the present paper, we focus on the WirelessHART technology, emphasizing its most important routing and scheduling aspects in order to guide both end users and the developers of new algorithms. Furthermore, we provide a detailed literature review of the newest routing and scheduling techniques forWirelessHART, discussing each of their features. These routing algorithms have been evaluated in terms of their objectives, metrics, the usage of theWirelessHART structures and validation method. In addition, the scheduling algorithms were also evaluated by metrics, validation, objectives and, in addition, by multiple superframe support, as well as by the redundancy method used. Moreover, this paper briefly presents some insights into the main WirelessHART simulation modules available, in order to provide viable test platforms for the routing and scheduling algorithms. Finally, some open issues in WirelessHART routing and scheduling algorithms are discussed.

Availability issues in wireless visual sensor networks.

Wireless visual sensor networks have been considered for a large set of monitoring applications related with surveillance, tracking and multipurpose visual monitoring. When sensors are deployed over a monitored field, permanent faults may happen during the network lifetime, reducing the monitoring quality or rendering parts or the entire network unavailable. In a different way from scalar sensor networks, camera-enabled sensors collect information following a directional sensing model, which changes the notions of vicinity and redundancy. Moreover, visual source nodes may have different relevancies for the applications, according to the monitoring requirements and cameras' poses. In this paper we discuss the most relevant availability issues related to wireless visual sensor networks, addressing availability evaluation and enhancement. Such discussions are valuable when designing, deploying and managing wireless visual sensor networks, bringing significant contributions to these networks.

Reliability and availability evaluation of Wireless Sensor Networks for industrial applications.

Wireless Sensor Networks (WSN) currently represent the best candidate to be adopted as the communication solution for the last mile connection in process control and monitoring applications in industrial environments. Most of these applications have stringent dependability (reliability and availability) requirements, as a system failure may result in economic losses, put people in danger or lead to environmental damages. Among the different type of faults that can lead to a system failure, permanent faults on network devices have a major impact. They can hamper communications over long periods of time and consequently disturb, or even disable, control algorithms. The lack of a structured approach enabling the evaluation of permanent faults, prevents system designers to optimize decisions that minimize these occurrences. In this work we propose a methodology based on an automatic generation of a fault tree to evaluate the reliability and availability of Wireless Sensor Networks, when permanent faults occur on network devices. The proposal supports any topology, different levels of redundancy, network reconfigurations, criticality of devices and arbitrary failure conditions. The proposed methodology is particularly suitable for the design and validation of Wireless Sensor Networks when trying to optimize its reliability and availability requirements.