The growing integration of sensors and embedded devices into distributed Internet of Things (IoT) systems has increased the demand for real-time data collection and processing solutions. However, the high volume and frequency of sensor data create challenges related to storage, transmission, and response latency, especially in resource-constrained environments. In this context, locally executed compression techniques, aligned with the Tiny Machine Learning (TinyML) paradigm, become differentiators for enabling embedded applications. Thus, this work proposes an autoencoder-based approach for efficiently compressing sensor data on edge devices. Three autoencoder variants (feed-forward, sparse, and contractive) are evaluated, combined with symmetric and asymmetric architectures, considering criteria such as compression ratio, information preservation, and embedded execution feasibility. For practical validation, a case study was conducted using vehicular data collected via the OBD-II interface, where the selected models were deployed on the OBDII Edge Freematics One+ device. The results show that the models could reduce data dimensionality with minimal information loss, maintain competitive performance on discriminative tasks, and exhibit inference times compatible with real-time applications. Autoencoders represent a viable neural compression solution for IoT environments, potentially applicable to various embedded sensing scenarios.

Kolmogorov–Arnold Networks (KANs) represent a promising machine learning architecture that leverages univariate functional decomposition to model complex phenomena using compact and interpretable structures. These characteristics make KANs especially attractive for deployment in TinyML environments, where memory, processing power, and energy consumption are strictly constrained. This paper evaluates the feasibility and trade-offs of using a KAN model to estimate the State of Charge (SoC) in electric vehicle batteries. We design a KAN tailored for embedded systems and compare its performance with a conventional Multilayer Perceptron (MLP) baseline under identical training and deployment conditions. Our evaluation includes predictive accuracy, training cost, model size, inference speed, and energy consumption on microcontrollers. Results show that the KAN model achieves a nearly ten times smaller memory footprint than the MLP (693 bytes vs. 6807 bytes) and maintains comparable energy consumption and inference speed across different embedded platforms. Although the MLP outperforms the KAN during training with faster convergence and lower energy requirements, the KAN demonstrates competitive predictive performance at inference time while significantly reducing deployment costs in terms of memory usage and energy efficiency at the edge. Furthermore, the KAN model produces symbolic mathematical expressions, offering direct interpretability and facilitating analytical validation — a critical advantage for embedded battery diagnostics and safety-critical applications.

The advancement of Industry 4.0 has intensified the demand for intelligent systems that can efficiently access and interpret technical information critical to industrial operations. However, recovering knowledge from extensive and complex technical documentation remains a significant challenge. This study examines the effectiveness of various Retrieval-Augmented Generation (RAG) strategies, combined with different Large Language Models (LLMs), in extracting and generating answers from industrial technical documents. A case study was conducted based on PROFIBUS, a widely adopted digital communication protocol in automation networks, with technical documents organized into categories for engineers and developers. Twenty questions of varying complexity were formulated, and responses were generated using three RAG strategies (Basic, Decomposition, and HyDE) combined with two LLMs (Gemma 3 and GPT-4o-mini). The outputs were compared against reference answers generated by the NotebookLM system and evaluated using automatic metrics, including ROUGE, METEOR, BERTScore, and MATTR. The results indicate that the Decomposition and HyDE strategies achieved superior semantic similarity scores when combined with more capable models. That model’s performance varied depending on the complexity of the document profile. These findings highlight the importance of tailored RAG strategies in enhancing intelligent information retrieval in industrial domains, which supports safer and more efficient operational environments.

The Internet of Things (IoT) generates massive multivariate time series data requiring real-time anomaly detection and correction for reliable monitoring. This challenges resource-constrained embedded devices due to conventional offline training and batch processing. To address this, we propose two algorithms derived from the TEDARLS framework: Multivariate Sequential TEDA (MST) and Multivariate Parallel TEDA (MPT). Derived from the TEDARLS framework, both enable on-device detection and correction of multivariate anomalies within TinyML constraints. A case study with real vehicular sensor data demonstrated low inference times and consistent embedded behavior. Supervised metrics were only assessed on synthetic data. MPT, though more sensitive, introduced greater signal distortions and required significantly longer processing times. Overall, MST demonstrated superior stability and suitability for real-time anomaly correction in resource-constrained IoT environments. This approach addresses an important gap in embedded analytics for IoT by enabling lightweight, accurate, and autonomous anomaly detection and correction at the edge.

The persistence of wildfires in natural landscapes calls for innovative early detection methods that leverage cutting-edge technologies. Traditional approaches, which rely solely on visual sensors or isolated devices, while valuable, often fall short in terms of accuracy, cost, scalability, and contextual adaptability. In response to these challenges, this paper introduces a novel fire detection system that integrates a sensor-based model with a dynamically triggered visual analysis module at edge devices. Central to our approach is a multi-sensor monitoring architecture that employs a TinyML classifier to continuously monitor environmental conditions under strict energy constraints. Upon detecting potential fire indicators, the system promptly activates a visual sensor that uses a camera platform to adjust its orientation based on the target position, capturing and analyzing images through a lightweight Convolutional Neural Network (CNN). This proposed system achieves an accuracy of up to 92%, while the quantized CNN models deliver an 83% reduction in inference time and a 74% and 70% decrease in peak RAM and Flash usage, respectively. Simulations also demonstrated that the system reduced the false-positive rate with minimal power increase.

This study addresses the challenges of dependability in Wireless Sensor Networks by proposing a Machine Learning-based approach using Convolutional Neural Networks for network planning for smart cities. Simulated scenarios were used to train the model, which predicts sensor placement and communication configurations to optimize coverage and availability. Results show significant improvements, including an average of 10.7% increase in dependability index and a rise in area coverage from 59% to 73% in 7-node networks, while reducing path failure rates by 27.6%. The method proves effective for enhancing WSN performance and adaptability in safety-critical applications.

The Internet of Things (IoT) has made it possible to include everyday objects in a connected network, allowing them to intelligently process data and respond to their environment. Thus, it is expected that those objects will gain an intelligent understanding of their environment and be able to process data more efficiently than before. Particularly, such edge computing paradigm has allowed the execution of inference methods on resource-constrained devices such as microcontrollers, significantly changing the way IoT applications have evolved in recent years. However, although this scenario has supported the development of Tiny Machine Learning (TinyML) approaches on such devices, there are still some challenges that require further investigation when optimizing data streaming on the edge. Therefore, this article proposes a new unsupervised TinyML regression technique based on the typicality and eccentricity of the samples to be processed. Moreover, the proposed technique also exploits a Recursive Least Squares (RLS) filter approach. Combining all these features, the proposed method uses similarities between samples to identify patterns when processing data streams, predicting outcomes based on these patterns. The results obtained through the extensive experimentation utilizing vehicular data streams were highly encouraging. The proposed algorithm was meticulously compared with the RLS algorithm and Convolutional Neural Networks (CNN). It exhibited significantly superior performance, with mean squared errors that were 4.68 and 12.02 times lower, respectively, compared to the aforementioned techniques.

The Internet of Things (IoT) paradigm has fostered several transformations in various industrial sectors, with important improvements in the automotive industry. Actually, the number of sensors and the computational power of modern vehicles have grown significantly, providing an opportunity for instrumentation, monitoring, and creation of increasingly efficient diagnostic algorithms. In fact, it is known that diagnosis is an essential requirement since the way of driving may have significant impacts in different contexts, such as traffic safety, fuel consumption, emissions, and maintenance, among others. Furthermore, solutions generally available in the literature for analyzing drivers’ behavior have focused on supervised offline learning models, fed with an entire dataset for training and testing. In this context, this paper proposes an approach for detecting drivers’ driving events, exploiting for that unsupervised online data flows and a specialized machine learning algorithm. The validation of the proposal was carried out with a case study in a real scenario with different conditions, which allowed the identification of daily driving operations. The results demonstrated the feasibility of the proposal as well as the identification of the different intended events.

Internet of things (IoT) is a trend which consists of providing connectivity for many consumer devices. Consequently, a considerable number of applications never seen before is emerging to make daily lifestyle more practical and connected through sharing, storing and processing personal data. When applied to industry, the same technologies can cause a real revolution in the manner how the information is treated. The value of the processed data is increasing fast and point to the fourth industrial revolution, or industry 4.0. This paper evaluates the performance of two ICT (Information and Communication Technology) protocols coming from the consumer world with an enormous potential to be used by industry. Experiments were performed to measure the round trip time using MQTT (Message Queuing Telemetry Transport) and WebSocket protocols with the exchange of data between one server in Italy and another in Brazil. The results can be analyzed for each possible application and different network paths. The average latency samples using each of the protocols were similar, defining the payload size of the requests, but changing according to the way.

In the last decade, the growth of the economic, automotive and technological sectors, has been notable. Alongside this growth, emerges the term ”Industry 4.0” which is used to represent the current Industrial Revolution. This revolution involves different areas: from manufacturing to healthcare. Industry 4.0 can create value during the entire product lifecycle, promoting customer feedback and having the entire product history throughout its life. 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 and so on. This would allow the prevention of critical issues and undesired behaviors as it prevents the automakers from losing contactwiththe vehicle after the purchase. 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 scanner the sensors available in vehicles, with the purpose of assisting in the management, prevention, and mitigation of different vehicular problems. An intercontinental experiment conducted in locations within Brazil and Italy show the feasibility of the platform and the potential to use preliminary results in order to improve the vehicle manufacturing.

Wireless sensor networks can provide visual information from the monitored field when sensor nodes are equipped with low-power cameras. In general, visual monitoring applications supported by sensing technology will have to address many challenging issues when visual information has to be transmitted over resource-constrained sensors. When addressing energy efficiency, sensing redundancy can be exploited to enlarge the network lifetime, whenever inactive sensors are used to replace faulty nodes. The monitoring of multiple targets may be optimized reducing the number of active visual sensors, but the required perspectives of the targets must be considered. In this paper we propose an algorithm to compute the minimum number of visual sensors that should be activated to cover all desired targets, especially addressing the particular problem when single nodes can view multiple targets at the same time. As different concurrent perspectives of the targets may be required, the proposed algorithm can bring significant results to wireless visual sensor network applications.

Visual monitoring in wireless sensor networks can provide valuable information of the monitored field, enriching surveillance and control applications. For those networks, however, some active visual sources may fail or run out of energy, potentially degrading the application monitoring quality. Visual sensors may be deployed to monitor a set of targets that are critical for the monitoring tasks of the application, demanding some level of redundancy to compensate sensor failures. In this context, it may be desired to know the probability of a specific target to be covered by at least one visual sensor along the network operation. We propose an approach for the availability assessment in wireless visual sensor networks for the specific case of target coverage, relating sensing redundancy to energy discharging and sensors disconnection. The proposed approach can then be used to predict coverage holes, directly benefiting critical monitoring applications.

The fault tree is a technique of high importance for reliability analysis in a industrial process. This paper describes an application that was developed for the automatic fault tree synthesis. In order to do so, the knowledge about industrial processes was structured through the development of a domain ontology. With this ontology is possible to represent the relationship between each component of the system and all its failure situations. Once the knowledge is structured in an ontology, an automatic procedure is executed to create a Fault Tree. The proposal was validated with a water tank system and allowed the construction of its fault tree quickly and efficiently.

As Wireless Sensor Networks have been employed to support critical monitoring applications, network availability has become a major design concern. In these networks, redundancy can be exploited to enhance the attainable availability level, where redundant sensors can replace faulty nodes. When camera-enabled sensors are deployed to retrieve visual information, the perception of redundancy changes considerably, since the redundancy of visual sensors depends on the monitoring requirements of the applications. In such context, characteristics as deployment density, viewing angle and sensing range are relevant when planning wireless sensor network applications, directly impacting in the number of redundant nodes. We propose an algorithm to select redundant nodes in Wireless Visual Sensor Networks, according to the application requirements. Moreover, we discuss how parameters of the deployed network can influence on the number of redundant nodes.

WirelessHART specification is emerging nowadays as a solution for the last mile connection in order to attend a trend on industrial networks that aims the usage of wireless technologies. Despite its high degree of applicability, a WirelessHART network faces some challenges. One of the most challenging problems are its reliability, energy consumption and the environment interference. Thus, representing a key component for the lower layers. In order to enable performance evaluations in a low cost and scalable way, computer simulations might be adopted. In this paper, we do a reliability and energy consumption evaluation over a faulty link industrial scenario using a new NS-3 module for the WirelessHART physical layer. This new model features includes an error model (Gilbert-Elliot), station positioning, signal attenuation and energy consumption. Furthermore, the module permits to configure each link with different fault probabilities, allowing to simulate more accurate scenarios.

There is a growing need to develop new tools to help end users in tasks related to the design, monitoring, maintenance and commissioning of industrial processes. The complexity of the industrial environment requires that these tools have flexible features in order to provide valuable data at the early planning and design phases. Furthermore, it is known that industrial processes have stringent requirements for dependability (reliability and availability), since failures can cause economic losses, environmental damages and danger to people. The lack of tools that enable the evaluation of faults in industrial processes could mitigate these problems. In this work we present a framework which is able to estimate the reliability and availability of industrial processes that can be modeled as a graph. The framework transforms the graph structure into a Fault Tree model in order to analyze their dependability. In order to validate the proposal we conduct a dependability study of an industrial process.

WirelessHART currently appears as a promising solution for the last mile connection in process control applications. Most of these applications have stringent dependability requirements, where a system failure may result in economic losses, or damage for human life or the environment. 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 communication over long periods of time and consequently they may disturb, or even disable, control algorithms with all resulting problems. In this work we perform a dependability evaluation of the best practices indicated by the HART Communication Foundation (HCF), when network devices are subject to permanent faults.

WirelessHART currently appears as a leading solution for interconnection of wireless devices in industrial process control applications. However, the lack of knowledge about the influence of transient faults in WirelessHART networks can lead to the choice of less reliable topologies. In this work, we propose a simulation model to evaluate WirelessHART networks in the presence of transient faults. We assume that these faults result from noisy environments that disturb communications between devices. The model was developed using the Stochastic Petri Net (SPN) formalism, supported by the Mbbius tool. For further developments, we target the development of an application that automates the assessment of a WirelessHART network in transient fault scenarios.

This work has the objective to present the first development results of a WirelessHART module for the ns-3. Our focus is the implementation of the Physical layer in order to provide the basis for the development of the superior layers such as MAC and Application. Thus, we presente an energy consumption model, a Gilbert/Elliot error model and an analysis for the currently avaliable ns-3 propagation loss models. For further development we mainly aim for the implementation of the time slot scheduler (Network Manager) and the development of an inter protocol simulation with mutual interference.

Wireless sensor network (WSN) is an emerging technology that targets multiple applications in the different environments. Its infrastructure is composed of a large number of sensor nodes with a limited physical capacity and low cost. The energy consumption must be as optimized as possible in order to extend its lifetime. The use of data compression techniques can be an advantage in the WSN context, once these techniques eliminate transmission of redundant information and consequently can be adopted to minimize the consumption of energy in the sensor nodes. WSN for monitoring applications can benefit from this technique as it may maximize the lifetime of batteries. The main motivation of this paper is to investigate the performance of a data compression algorithm for WSN in the context of monitoring applications. To validate the proposal, simulation experiments have been performed using the Network Simulator (NS-2) tool.