Journal of Ambient Intelligence and Smart Environments

The rapid advances in the Internet of Things (IoT) and the increment of its users have opened a door to put into service IoT in healthcare, known as the Internet of Medical Things (IoMT), which consists of cloud, fog, and edge computing. IoMT has shown to be a flexible framework to remove traditional healthcare limits through utilizing technology/techniques and innovative digital devices to monitor patients’ conditions and to address medical service problems such as inaccuracy and long response time. It also could be helpful within critical and unexpected circumstances like epidemic diseases. Fortunately, some methods, including Edge Computing, were introduced to improve the healthcare system, make it more effective and solve problems. In this work, a Systematic Literature Review (SLR) was used to study Edge Computing solutions in healthcare, evaluate efficient therapeutic approaches, and demonstrate key factors that have not been considered in previous studies.

With the vigorous development of information technology, the applications of the Internet of Things (IoT) have become increasingly common in recent years. Robot vacuum has become a popular and representative product in smart homes. This study proposed a hybrid fuzzy multi-criteria decision-making (MCDM) model that applied fuzzy analytic network process (FANP) and decision-making trial and evaluation laboratory (DEMATEL) to analyze the critical factors evaluated by users when adopting a robot vacuum. It was found that the top two dimensions in order are “epistemic value” and “functional value”; and the top five factors in order are “novelty”, “exploratory”, “family information infrastructure”, “family consensus”, and “reliability”. Significant influential and affected factors were identified. Gender differences in decision-making factors are also discussed.

Task allocation is a vital challenge in a multi-robot environment. A hybrid fuzzy response threshold-based method is proposed to address the problem of task allocation in a heterogeneous mobile robot environment. The method follows a distributed task allocation approach where every robot chooses its task and performs it, resulting in concurrent execution. The algorithm uses a fuzzy inference system to determine the capability of the robot to carry out a task. Then, the robot employs the response threshold model, utilizing the obtained capability to decide on the task to complete. The objective here is to maximize the tasks completed with the resources available while balancing the affinity with which the task is done. The proposed algorithm is initially applied to the static scenario where there is no failure among the mobile robots. The algorithm is then improved to run in the dynamic scenario to study the effect on the allocation. The proposed algorithm is empirically evaluated in simulation for multiple runs under different environment instances. The results show a good increase in tasks performed successfully across all the instances in static and dynamic scenarios. The proposed algorithms are validated using FireBird V mobile robots in an experimental environment.

Smart homes integrate several sensors to facilitate information exchange and the execution of tasks. In addition, with the development of the Internet of Things (IoT) platforms, the control of appliances and remote devices has become possible. This sensor collects data in real time to closely monitor the devices of a user’s household. The present study employs a machine learning methodology to perform a global analysis of energy consumption and efficiency in smart homes. In This work we propose two advanced ensemble models to improve the performance of energy consumption in smart homes, the first one is a voting ensemble model based on a ranking weight averaging that combines following basic machine learning techniques: decision tree (DT), random forest (RF), and eXtreme Gradient Boosting (XGB). The second one is the stacking ensemble model in which the basic models (DT-RF-XGB) are combined through stacked generalization, then uses a secondary layer model or meta-learner (RF) to provide output prediction. The findings obtained show that the proposed ensemble model based on DT-RF-XGB using stacking technique surpasses all other basic algorithms with R² around 0.9825.

Technological innovations in Internet of Things (IoT) have resulted in smart agricultural solutions such as a remotely monitored Aquaponics system and a wireless sensor network (WSN) of such systems (nodes). IoT enables continuous sensing of temperature and pH data at each node of the WSN, which is periodically transmitted to a remote fusion centre. In this regard, the data matrices acquired at the fusion centre often suffer from data vacancies and missing data problems, owing to typical wireless multipath fading environment, sensor malfunctions and node failures. This paper explores the applicability of different matrix completion approaches for missing data reconstruction. Specifically, the performance of baseline predictor, correlation based approaches such as baseline predictor with temporal model, k-nearest neighbors (kNN) and low rank based approaches such as Sparsity Regularized Singular Value Decomposition (SRSVD) and Augmented Lagrangian Sparsity Regularized Matrix Factorization (ALSRMF) have been explored. Reliable temperature and pH data for 19 independent acquisition hours with 60 samples per hour are acquired at the fusion centre via Ultra High Frequency (UHF) transmission at 470 MHz and suitable pre-processing. Simulating different data integrity scenarios, the reconstruction error plots from each of these matrix completion approaches is extracted. A hybrid of kNN and baseline predictor with temporal model rendered a Mean Absolute Percentage Error (MAPE) of 1.75% for temperature and 0.86% for pH, at 0.5 data integrity. Further, with ALSRMF, which exploits the low rank constraint, the error reduced to 1.25% for temperature and 0.7% for pH, thus substantiating a promising approach for Aquaponics system data reconstruction.

Human activity recognition (HAR) is a crucial area of research in human-computer interaction. Despite previous efforts in this field, there is still a need for more accurate and robust methods that can handle time-series data from different sensors. In this study, we propose a novel method that generates an image using wavelet transform to extract time-frequency features of the recorded signal. Our method employs convolutional neural networks (CNNs) for feature extraction and activity recognition, and a new loss function that produces denser representations for samples, improving the model’s generalization on unseen samples. To evaluate the effectiveness of our proposed method, we conducted experiments on multiple publicly available data sets. Our results demonstrate that our method outperforms previous methods in terms of activity classification accuracy. Specifically, our method achieves higher accuracy rates and demonstrates improved robustness in real-world settings. Overall, our proposed method addresses the research gap of accurate and robust activity recognition from time-series data recorded from different sensors. Our findings have the potential to improve the accuracy and robustness of human activity recognition systems in real-world applications.

The manual sorting of recyclable garbage has caused several issues such as the wastage of human resources and low resource utilization. To solve this problem, an improved Single Shot Multibox Detector (SSD) deep learning approach has been developed for recyclable garbage detection. To reduce the number of parameters and make the model easier to deploy and apply, a lightweight network called RepVGG has been chosen to replace the VGG16 network in the SSD. Additionally, the auxiliary convolutional layer structure of the SSD has been modified to further reduce the number of parameters. Additionally, the SK module has been integrated to adaptively adjust the size of the receptive field and enhance the detection accuracy. Experimental results of Waste Classification data set from Kaggle website have demonstrated that the improved SSD model has better detection accuracy and real-time performance, with an accuracy of 95.23%, which is 4.33 percentage points higher than the original SSD, and a detection speed of up to 64 FPS. This algorithm can be better applied in industry.

Nowadays, the object’s volume is essential for monitoring any scene. Technological equipment is evolving, and mobile devices and other devices embed high-resolution cameras. The high-resolution cameras open a window for different research studies, where the volume measurement is vital for different areas. This study aims to identify image processing techniques for measuring the object’s volume. Thus, a systematic review was performed with a Natural Language Processing (NLP)-based framework for identifying studies between 2010 and 2023 related to the measurement of object volume. As a result of this search, this paper reviewed and analyzed 25 studies, verifying that different computer vision methods accurately handle object recognition. Additionally, an evaluation of the databases presented by the studies above is performed to consider further the design of a new approach to infer the volume of objects from an image.