Environmental monitoring is a practical application where a Wireless Sensor Network (WSN) may be utilized effectively. However, the energy consumption issues have become a major concern in using a WSN, particularly in remote locations without readily accessible electrical power supply. In general, the activities of data transmission among sensor nodes and the gateway can be a significant fraction of the total energy consumption within a WSN. Hence, reducing the number and the duration of transmissions as much as possible while maintaining a high level of data accuracy can be an effective strategy for saving energy. To achieve this objective, a Least Mean Square (LMS) filter is used for a dual prediction scheme, in this paper. The dual prediction scheme is data quality-based, allowing both the sensor nodes and the gateway to predict the data simultaneously. Only when the error between the predicted data and the real sensed data exceeds a pre-defined threshold, the sensor nodes will send the sensed data to the gateway/another node and consequently will update the coefficients of the filter. It is observed that, with this scheme, the total number of transmissions and their overall duration can be effectively reduced, and therefore, further energy savings can be realized. With the developed methodologies, at least 62.3% of the total energy for data transmission could be saved while achieving a 93.1% prediction accuracy.

In this paper, the problem of the cross-domain fault diagnosis of rotating machinery is considered. In a practical setting of this approach, the operating platform of the machine may have a different setup and conditions compared to the experimental platform that is used to collect the training data. This can lead to significant data variations, specifically, domain shifts. Conventional data-driven approaches are known to adapt poorly to these domain shifts, resulting in a significant drop in the diagnosis accuracy when the pre-trained model is applied in the actual operating situation. In this paper, an unsupervised domain adaptation approach is developed to mitigate the domain shifts between the data gathered from the experimental platform (the source domain) and the operating platform (the target domain) by aligning the features extracted from the two data domains. The mutual information between the target feature space and the entire feature space is maximized to improve the knowledge transferability of the labeled data in the source domain. Furthermore, the feature-level discrepancy between the two domains is minimized to further improve diagnosis accuracy. The experiments using public datasets and real-world adaptation scenarios demonstrate the feasibility and the superior performance of the proposed method.

Power management is crucial in the monitoring of a remote environment, especially when long-term monitoring is needed. Renewable energy sources such as solar and wind may be harvested to sustain a monitoring system. However, without proper power management, equipment within the monitoring system may become nonfunctional and, as a consequence, the data or events captured during the monitoring process will become inaccurate as well. This paper develops and applies a novel adaptive sampling algorithm for power management in the automated monitoring of the quality of water in an extensive and remote aquatic environment. Based on the data collected on line using sensor nodes, a data-driven adaptive sampling algorithm (DDASA) is developed for improving the power efficiency while ensuring the accuracy of sampled data. The developed algorithm is evaluated using two distinct key parameters, which are dissolved oxygen (DO) and turbidity. It is found that by dynamically changing the sampling frequency, the battery lifetime can be effectively prolonged while maintaining a required level of sampling accuracy. According to the simulation results, compared to a fixed sampling rate, approximately 30.66% of the battery energy can be saved for three months of continuous water quality monitoring. Using the same dataset to compare with a traditional adaptive sampling algorithm (ASA), while achieving around the same Normalized Mean Error (NME), DDASA is superior in saving 5.31% more battery energy.

An Internet of Things (IoT) platform with capabilities of sensing, data processing, and wireless communication has been deployed to support remote aquatic environmental monitoring. In this paper, the design and development of an IoT platform with multiple Mobile Sensor Nodes (MSN) for the spatiotemporal quality evaluation of surface water is presented. A survey planner is proposed to distribute the Sampling Locations of Interest (SLoIs) over the study area and generate paths for MSNs to visit the SLoIs, given the limited energy and time budgets. The SLoIs are chosen based on a cellular decomposition that is composed of uniform hexagonal cells. They are visited by the MSNs along a path ring generated by a planning approach that uses a spanning tree. For quality evaluation, an Online Water Quality Index (OLWQI) is developed to interpret the large quantities of online measurements. The index formulations are modified by a state-of-the-art index, the CCME WQI, which has been developed by the Canadian Council of Ministers of Environment (CCME) for off-line indexing. The proposed index has demonstrated effective and reliable performance in online indexing a large volume of measurements of water quality parameters. The IoT platform is deployed in the field, and its performance is demonstrated and discussed in this paper.

This paper concerns the deployment problem of wireless sensor networks (WSNs) with mobile robotic sensor nodes for spatiotemporal monitoring. The proposed approach, deep reinforced learning tree (DRLT), utilizes deep reinforcement learning (DRL) to improve the efficiency of searching the most informative sampling locations. The parameterized sampling locations in an infinite horizon space are modeled according to their spatiotemporal correlations and subject to various constraints, including field estimation error and information gain. And the model-based information gain can be calculated efficiently over an infinite horizon. In this manner, the effectiveness of the sampling locations is learned through DRLT during the exploration by the robotic sensors. Then DRLT can instruct the robotic sensors to avoid unnecessary sampling locations in future iterations. Also, it is proved in this paper that the proposed algorithm is capable of searching for the near-optimal sampling locations effectively and guaranteeing a minimum field estimation error. Simulation based on national oceanic and atmospheric administration (NOAA) datasets is presented, which demonstrates the significant enhancements made by the proposed algorithm. Compared with the traditional approaches, such as the information theory-based greedy approach or random sampling, the proposed method shows a superior performance with regard to both estimation error and planning efficiency.

This paper addresses the problem of optimizing sensor deployment locations of a wireless sensor network to reconstruct and also predict a spatiotemporal signal. Traditional sensor deployment optimization approaches only selects deployment locations and provide the feature in the sampled area. Hence, if the sampling resources are limited, the deployed sensors may not be able to provide all features in the field of interest (input space). To solve this issue, a deep learning framework is developed to reconstruct and predict the entire spatiotemporal signal from a limited amount of observations. In the beginning, the proposed approach optimizes sampling locations to retrieve sufficient features for reconstruction and prediction from historical data. Then, a spatiotemporal autoencoder is used to capture the nonlinear mappings from the in-situ measurements to the entire spatiotemporal signal. A simulation is conducted using global climate datasets from the National Oceanic and Atmospheric Administration, to implement and validate the developed methodology. The results demonstrate a significant improvement made by the proposed algorithm. Specifically, compared to traditional approaches, the proposed method provides superior performance in terms of both reconstruction error and spatial prediction robustness.