Smart Environments Research Articles

DEEP LEARNING FRAMEWORK FOR HUMAN ACTIVITY DETECTION BASED ON WIRELESS SENSORS.

Recognizing human athletic behaviors has grown in importance as a component of analysis during the last few years. For upcoming training, FPGA-based recognition has been built to perfectly predict human actions. Intelligent environments where human athletic movements may be automatically recognized are required to ensure the viability of such long-term athletic training monitoring systems. The system discussed here employs an Artificial Neural Network (ANN) based Field-Programmable Gate Array (FPGA). In order to infer human behavior from the data collected by a smart environment, machine learning algorithms must first be trained on annotated datasets. To effectively extract characteristics from unprocessed data and develop a machine learning model for anticipating an individual's movement, conventional signal processing techniques and domain knowledge are required. The purpose of this work is to show how a hybrid deep learning model may be applied to recognize human behavior. Convolutional neural networks and continuous neural networks, for example, are deep learning techniques that will extract the features and help with categorization. The proposed model forecasts human activity using wireless sensor data mining datasets. Accuracy, training loss, testing loss, the confusion matrix, and other metrics have all been used to evaluate the model's performance.

A Decade of Innovation Ecosystem Development: Bibliometric Review of Scopus Database

The paper focuses on the bibliometric review of the Scopus database in the field of innovation ecosystem development, aiming to reveal the key trends in this fast-growing area of interdisciplinary research in terms of different quantitative and qualitative parameters. The bibliometric analysis followed PRISMA protocol guidelines for finding and extracting relevant scholarly papers based on the selected national, institutional, demographic, and scientific variables. The PRISMA procedure resulted in 401 selected open-access articles published on the topic of innovation ecosystem development from 2013 to 2023. The key findings indicate that research on innovation ecosystem development has seen late growth, which is a sign of still underexplored fields for potential pioneers. Researchers and institutions from the Nordic countries and the UK are most active as far as publications on innovation ecosystem development are concerned. However, researchers from the USA, Australia, and the UK dominate the citation records. Research collaborations help increase productivity and citation levels. The most cited articles fall into 4 clusters based on citations: innovation, smart tourism, digitalization, and entrepreneurship. Future research synergies can also be envisaged with the domains of digitalization, sustainable development, and the smart environment.

Ubiquitous multi-occupant detection in smart environments

Recent advancements in ubiquitous computing have emphasized the need for privacy-preserving occupancy detection in smart environments to enhance security. This work presents a novel occupancy detection solution utilizing privacy-aware sensing technologies. The solution analyzes time-series data to detect not only occupancy as a binary problem, but also determines whether one or multiple individuals are present in an indoor environment. On three real-world datasets, our models outperformed various state-of-the-art algorithms, achieving F1-scores up to 94.91% in single-occupancy detection and a macro F1-score of 91.55% in multi-occupancy detection. This makes our approach a promising solution for improving security in smart environments.

MultiBEATS: Blocks of eigenvalues algorithm for multivariate time series dimensionality reduction

Multivariate Time Series are sequences of observations taken from multiple sources. The proliferation of environments in which data is collected by means of sensors and the adoption of data-based services reveals the importance of their efficient analysis. In smart environments, the analysis of the raw Multivariate Time Series is cumbersome. The algorithms are heavy, slow, and fail to extract all the knowledge available. In that sense, representation techniques reduce the data points but maintain the inner patterns so that the information present in data persists. There exist several approaches to represent Univariate Time Series, but for the multivariate case they are scarce. We have adapted the existent univariate representation algorithm BEATS, which is based on Discrete Cosine transformation and the extraction of eigenvalues to multivariate scenarios and we have called it MultiBEATS. Our method allows flexibility with regard to data reduction and does not require labelled data. To investigate the efficacy of MultiBEATS, we selected 8 open multivariate time series datasets and compared the running time and accuracy of their classification using raw data and MultiBEATS transformed data. We have also compared our results with the representation algorithm Seq2VAR, based on autoencoders, for having a baseline. The MultiBEATS transformation reduces the execution time of classification algorithms, has improved the results in accuracy in 2 of the datasets and matches a third one. With regards to the baseline, MultiBEATS is faster than Seq2VAR and more accurate in 7 out of the 8 datasets. In conclusion, MultiBEATS is a promising approach for efficiently reducing multivariate time series data, therefore helping decision-making in dynamic smart environments.

3D Human Pose Estimation from multi-view thermal vision sensors

Human Pose Estimation from images allows the recognition of key daily activity patterns in Smart Environments. Current State-of-the-art (SOTA) 3D pose estimators are built on visible spectrum images, which can lead to privacy concerns in Ambient Assisted Living solutions. Thermal Vision sensors are being deployed in these environments, as they preserve privacy and operate in low brightness conditions. Furthermore, multi-view setups provide the most accurate 3D pose estimation, as the occlusion problem is overcome by having images from different perspectives. Nevertheless, no solutions in the literature use thermal vision sensors following a multi-view scheme.In this work, a multi-view setup consisting of low-cost devices is deployed in the Smart Home of the University of Almería. Thermal and visible images are paired using homography, and SOTA solutions such as YOLOv3 and Blazepose are used to annotate the bounding box and 2D pose in the thermal images. ThermalYOLO is built by fine-tuning YOLOv3 and outperforms YOLOv3 by 5% in bounding box recognition and by 1% in IoU value. Furthermore, InceptionResNetV2 is found as the most appropriate architecture for 2D pose estimation.Finally, a 3D pose estimator was built comparing input approaches and convolutional architectures. Results show that the most appropriate architecture is having three single-channel thermal images processed by independent convolutional backbones (ResNet50 in this case). After these, the output is fused with the 2D poses. The resulting convolutional neural network shows excellent behaviour when having occlusions, outperforming single-view SOTA approaches in the visible spectrum.

How to manage massive spatiotemporal dataset from stationary and non-stationary sensors in commercial DBMS?

AbstractThe growing diffusion of the latest information and communication technologies in different contexts allowed the constitution of enormous sensing networks that form the underlying texture of smart environments. The amount and the speed at which these environments produce and consume data are starting to challenge current spatial data management technologies. In this work, we report on our experience handling real-world spatiotemporal datasets: a stationary dataset referring to the parking monitoring system and a non-stationary dataset referring to a train-mounted railway monitoring system. In particular, we present the results of an empirical comparison of the retrieval performances achieved by three different off-the-shelf settings to manage spatiotemporal data, namely the well-established combination of PostgreSQL + PostGIS with standard indexing, a clustered version of the same setup, and then a combination of the basic setup with Timescale, a storage extension specialized in handling temporal data. Since the non-stationary dataset has put much pressure on the configurations above, we furtherly investigated the advantages achievable by combining the TSMS setup with state-of-the-art indexing techniques. Results showed that the standard indexing is by far outperformed by the other solutions, which have different trade-offs. This experience may help researchers and practitioners facing similar problems managing these types of data.

Impact of Peer Review on Learning Performance in a Smart Classroom Teaching Environment

With the rapid progress of educational technology, smart classrooms have gradually been widely applied, aiming to provide students with more efficient and innovative learning experiences. As a non-traditional assessment method, peer review has attracted widespread attention in this context, and its role in the learning process of students is increasingly prominent. However, there are still disputes and deficiencies regarding its specific applications and benefits in smart classroom environments. This study aimed to dig into the peer review standard and its score prediction in a smart classroom environment and evaluate the specific impact of peer review on learning performance. It is expected that this study can provide educators with a more accurate and practical peer review method, thereby optimizing the teaching and assessment modes of smart classrooms.

The influence of material and spatial size on indoor environment preference evaluation

ABSTRACT The effect the material-built environments exert on people’s perception is vital for people’s wellbeing. However, the correlation between the material and the space is under-researched. The development of smart building technologies brings opportunities and challenges to explore this problem, no matter for the material type or for the study method. In this research, 18 indoor environments in two sizes covered by nine materials in VR are used to examine the influence of the materials on people’s perceptions. This research uses a preference survey as the primary data gathering tool. Statistical tests used are Kruskal-Wallis and Mann-Whitney. The results reveal that there is a significant difference in the preference level among the simple cubic indoor environments covered with different materials. The results also reveal that the change in size appears to influence the material preference level following a certain rule, but it can induce a significant effect on the preference level of only several specific materials. This study furthermore highlights that the results may be useful to extend the knowledge that designers and architects have of the perception to indoor environments and may propose a new idea for generating or improving a smart indoor environment.

Information Communication Technology (ICT) and Education

COVID-19 has accelerated the shift to blended or fully online learning environments, enforcing educational institutions to embrace technology and offer their students an online or at least blended learning experience [...]

UIoTN‐PMSE: Ubiquitous IoT network‐based predictive modeling in smart environment

SummaryWe proposed a three‐stage intrusion detection system that utilizes a predictive machine learning model to identify and mitigate attacks on ubiquitous network. In the first stage, we applied Apriori‐enabled Association Rule Mining (AARM) feature selection with support vector machine (SVM) for classification of flow of network. Second, we proposed ensemble learning‐based AARM model (PAEL) for behavior analysis. Finally, for classification of multi‐task labels, we proposed swarm bat optimization‐based PAEL model. The trained model is applied to edge and fog computing devices to obtain lower resource utilization and improve the efficiency of the system. The intrusion detection process is performed in three stages: (i) at the edge devices, where abnormal data from network traffic from IoT devices were identified, (ii) the abnormal data sample is sent to fog computing deivce to confirm the attacks and abnormalities, (iii) final identified data sample is sent to cloud server. At cloud, proposed predictive machine learning (ML)‐based generalized weight sum‐enabled ensemble learning (PML‐GWEL) model is trained on sample data, including new detected samples, to continually improve its accuracy. Once the model is trained, it is published to all nodes in the network to update their primary detector models and clear out any outdated pre‐detector models. This process helps to reduce the hardware resources used by the pre‐detector models and improve the overall efficiency of the system. The proposed model is compared with other existing techniques.

A vision-based multi-cues approach for individual students’ and overall class engagement monitoring in smart classroom environments

A vision-based multi-cues approach for individual students’ and overall class engagement monitoring in smart classroom environments

Designs and practices using generative AI for sustainable student discourse and knowledge creation

Utilizing generative artificial intelligence, especially the more popularly used Generative Pre-trained Transformer (GPT) architecture, has made it possible to employ AI in ways that were previously not possible with conventional assessment and evaluation technologies for learning. As educational use cases and academic studies become increasingly prevalent, it is critical for education stakeholders to discuss design considerations and ideals that are key in supporting and augmenting learning via quality classroom discourse that sets the climate for student learning and thinking, and teachers’ transmission of expectations. In this paper, we seek to address how emergent technological advancements such as GPT, can be considered and utilized in designs that are consistent with the ideals of sustainable student discourse and knowledge creation. We showcase contemporary exemplars of possible designs and practices that are based on the pedagogy of knowledge building, with recent illustrations of how GPT may be utilized to sustain students’ knowledge building discourse. We also examine the potential effects and repercussions of technological utilization and misuse, along with insights into GPT’s role in supporting and enhancing knowledge building practices. We anticipate that the findings, through our exploration of designs and practices for knowledge creation, will be able to resonate with a broader audience and instigate meaningful change on issues of teaching and learning within smart learning environments.

Hybrid intrusion detection model for Internet of Things (IoT) network environment

Technology development brought numerous lifestyle changes. People move around with smart gadgets and devices in the home, work environment, and familiar places. The Internet acts as a backbone for all applications and connecting multiple devices to set up a smart environment is technically termed as IoT (Internet of Things). The feature merits of IoT are explored in numerous fields from simple psychical data measurement to complex trajectory data measurement. Where the place is inaccessible to humans, IoT devices are used to analyze the region. Though IoT provides numerous benefits, due to its size and energy limitations, it faces security and privacy issues. Intrusions in IoT networks have become common due to these limitations and various intrusion detection methods are introduced in the past decade. Existing learning-based methods lag in performance while detecting multiple attacks. Conventional detection models could not be able to detect the intrusion type in detail. The diverse IoT network data has several types of high dimensional features which could not be effectively processed by the conventional methods while detecting intrusions. Recently improvements in learning strategies proved the performance of deep learning models in intrusion detection systems. However, detecting multiple attacks using a single deep learning model is quite complex. Thus, in this research a multi deep learning model is presented to detect multiple attacks. The initial intrusion features are extracted through the AlexNet, and then essential features are selected through bidirectional LSTM. Finally, the selected features are classified using the decision tree C5.0 algorithm to attain better detection accuracy. Proposed model experimentations include benchmark NSL-KDD dataset to verify performances and compared the results with existing IDSs based on DeepNet, Multi-CNN, Auto Encoder, Gaussian mixture, Generative adversarial Network, and Convolutional Neural Network models. The proposed model attained maximum detection accuracy of 98.8% over conventional methods. Overall, an average of 15% improved detection performance is attained by the proposed model in detecting several types of intrusions in the IoT network.

Strengthening Privacy and Data Security in Biomedical Microelectromechanical Systems by IoT Communication Security and Protection in Smart Healthcare

Biomedical Microelectromechanical Systems (BioMEMS) serve as a crucial catalyst in enhancing IoT communication security and safeguarding smart healthcare systems. Situated at the nexus of advanced technology and healthcare, BioMEMS are instrumental in pioneering personalized diagnostics, monitoring, and therapeutic applications. Nonetheless, this integration brings forth a complex array of security and privacy challenges intrinsic to IoT communications within smart healthcare ecosystems, demanding comprehensive scrutiny. In this manuscript, we embark on an extensive analysis of the intricate security terrain associated with IoT communications in the realm of BioMEMS, addressing a spectrum of vulnerabilities that spans cyber threats, data manipulation, and interception of communications. The integration of real-world case studies serves to illuminate the direct repercussions of security breaches within smart healthcare systems, highlighting the imperative to safeguard both patient safety and the integrity of medical data. We delve into a suite of security solutions, encompassing rigorous authentication processes, data encryption, designs resistant to attacks, and continuous monitoring mechanisms, all tailored to fortify BioMEMS in the face of ever-evolving threats within smart healthcare environments. Furthermore, the paper underscores the vital role of ethical and regulatory considerations, emphasizing the need to uphold patient autonomy, ensure the confidentiality of data, and maintain equitable access to healthcare in the context of IoT communication security. Looking forward, we explore the impending landscape of BioMEMS security as it intertwines with emerging technologies such as AI-driven diagnostics, quantum computing, and genomic integration, anticipating potential challenges and strategizing for the future. In doing so, this paper highlights the paramount importance of adopting an integrated approach that seamlessly blends technological innovation, ethical foresight, and collaborative ingenuity, thereby steering BioMEMS towards a secure and resilient future within smart healthcare systems, in the ambit of IoT communication security and protection.

IoT Resources and Their Practical Application, A Comprehensive Study

The Internet of Things (IoT) has become a paradigm shifter, connecting an enormous number of smart devices and facilitating seamless data exchange for a diverse array of applications. The availability and effective use of the IoT ecosystem's resources are key factors in determining how its practical applications will develop as they mature. The IoT resources and their practical application across several areas are thoroughly explored in this paper. The paper begins by classifying and describing the various sensor types, their applications in various fields, and IoT resources, highlighting their contributions to real-time data collection, processing, and transmission. It then goes on to demonstrate a wide range of real-world uses for these resources, such as smart cities, education, agriculture, business, healthcare, environment monitoring, transportation, and industrial automation. However, utilizing IoT resources effectively is not without difficulties. Critical difficulties such as resource allocation, scalability, security, interoperability, and privacy concerns are identified and discussed in the paper. Furthermore, the paper also highlights future directions and emerging trends in IoT resource management, including edge computing, cloud computing, human machine integration, and compatibility with other systems. These developments aim to increase the dependability of IoT applications in diverse settings and optimize resource allocation. This paper's conclusion highlights the crucial role that IoT resources play in advancing real-world applications across a variety of areas. Researchers, practitioners, policymakers, and other stakeholders may collaborate together to effectively leverage the full potential of IoT resources to build intelligent, effective ecosystems that meet the needs of contemporary society by solving difficulties and utilizing developing trends.

Bi-Level Optimization Model for DERs Dispatch Based on an Improved Harmony Searching Algorithm in a Smart Grid

To satisfy the large-scale grid-connected demands of distributed energy resources (DERs), a bi-level optimization model is proposed for private DER dispatch managed by a virtual power plant (VPP) in a smart grid environment in this paper. The optimization models of the upper and lower layers serve as the planning layer and the operational layer, respectively, wherein the upper optimization model is mainly aimed at the optimization allocation of DERs under an unchanged network topology, and the lower-level optimization model is mainly responsible for the optimization of the distributed network topology from the optimization results of the upper layer. An improved harmony searching algorithm (HSA) is then applied to resolve the model through continuous iterations and communications between the upper and lower levels until an expected aim is achieved, which can satisfy the aim demand of the upper and the lower levels at the same time. Finally, six optimization scenarios are planned and selected in order to evaluate the performance of the bi-level optimization strategy in the IEEE33 bus distribution system and the 69-bus distribution system, and the optimized results show that the proposed method can effectively improve the grid-connected capacity of DERs and their system performance, and are more flexible in facing a growing ambient smart environment.

Smart environment practices in Kashiwa-No-Ha (Japan) smart city

Abstract The smart city is a modernization effort that enables cities to use their resources more effectively and to provide better services to their residents. In this study, the smart practices of Kashiwanoha City, Japan, one of the best smart city examples in the world, were examined. The study showed that smart environment applications are designed to improve the quality of life of the citizens and ensure the sustainability of the city. The smart environment applications in Kashiwa-no-ha Smart City include smart mobility, smart technology, smart governance, smart economy, smart living, and smart environment based on transport, energy, and environment. The integration of blockchain technology is a potential solution to address the challenges of urbanisation in smart sustainable city development. The development of Kashiwa-no-ha Smart City is an example of how cities are linking their technological policies and development plans to achieve development while linking the objectives and dimensions of sustainable development to provide the best possible benefit from the available technologies. The most important factor behind Kashiwa-no-ha‘s success is the cooperation of the public and private sectors and universities. For successful smart city practices, it has been proposed to establish an independent organisation similar to the Urban Design Centre Kashiwanoha (UDCK) with a separate budget and decision-making power.

Monitoring soil elements for irrigation management using Internet of Things (IoT) sensors

AbstractSmart farming focuses on three crucial areas: soil quality, weather conditions, and crop health. Because of the significant impact of weather and irrigation factors, many smart farming solutions are tightly integrated with smart environment and smart water (pollution, turbidity, nutrients) systems for a holistic approach. The Internet of Things (IoT) is revolutionizing soil monitoring, giving farmers and growers the means to optimize yields, minimize disease risks, and rationalize resource use. Thanks to IoT sensors, it is possible to measure various parameters such as soil temperature, NPK (nitrogen, phosphorus, potassium) levels, and soil moisture. The information collected by these IoT sensors is then transmitted to a central point or stored in the cloud server for in‐depth analysis, visual representation, and trend identification. The application of the IoT in agriculture is widely referred to as smart farming. The IoT is an essential pillar of precision agriculture, enabling more targeted and efficient management of resources. Smart sensor systems provide more information about water and crop water requirements. This information can be used to mechanically manage the water supply system and help farmers optimize their irrigation system. The information acquired by the first sensors is transferred to the cloud server. This article proposes and evaluates the concept of remote sensing systems. Three sites were selected to test the IoT system. At each site, five sensors were employed, specifically designed for soil and irrigation water analysis. The sensors dedicated to soil pH, moisture, and NPK demonstrated commendable accuracy; equally, the irrigation water pH and turbidity sensors exhibited notably precise measurements. The results were noteworthy, with an R2 value surpassing 90% across all sensor measurements. These IoT sensors could facilitate effective crop monitoring without excessive expenditures and serve to safeguard groundwater from nutrient contamination.

Review of routing protocol for low power and lossy network in the internet of things

The growth of internet of things (IoT) devices and apps has caused disruptions in key sectors such as smart environment, healthcare, and mission-critical tool. Security of IoT devices, networks, and infrastructure is the biggest challenge. RPL, a routing protocol optimised for low-power and lossy networks, is applied by the IoT to offer seamless information transfer among the myriad of interconnected entities. So, the paper delves into the nuts and bolts of RPL, covering its protocol settings, trickle timer, and objective function, as well as the IoT’s architecture and applications. This article also provides a framework for categorising past research on the effects of RPL-IoT on energy consumption, transportation, service quality, traffic congestion, and safety. The future of energy efficiency, congestion, objective function, mobility, stability, security, collaborative intrusion detection systems, active learning, and Key Management are all considered in this analysis.

Generating voice user interfaces from web sites

ABSTRACT Virtual assistants allow users to interact with apps using a conversational mode. A particular kind of device supporting virtual assistants is the smart speaker, which allows end users to access contents and services, perform searches and command smart environments via voice interaction. These devices are gaining relevance among users, and the number of apps available for giving them more conversational capabilities grows constantly. However, there is a gap between what can be achieved using these devices and the information available on the web, i.e. most web apps don't have one smart speaker app counterpart. In this work, we present a two-step approach to defining conversational interfaces for virtual assistants based on existing web sites. During the first step, the user creates web content blocks, which describe target contents and the strategy for retrieving them. In the second step, the user uses web content blocks to specify the conversational interface. Although the approach could be considered generic enough for any kind of conversational interface, we focus our research on voice user interfaces, considering smart speaker apps as a major target. With our approach, apps based on voice user interfaces for smart speakers may be developed in a no-code manner. We describe and illustrate the approach by presenting usage examples and an evaluation using the Alexa service and an Amazon Echo device. The evaluation shows promising results for two case studies where there is no significant difference between the user experience when comparing solutions developed using the Alexa SDK versus apps using our approach.