Large-scale Sensor Systems Research Articles

Multimodal reaching-position prediction for ADL support using neural networks

This study aimed to develop daily living support robots for patients with hemiplegia and the elderly. To support the daily living activities using robots in ordinary households without imposing physical and mental burdens on users, the system must detect the actions of the user and move appropriately according to their motions. We propose a reaching-position prediction scheme that targets the motion of lifting the upper arm, which is burdensome for patients with hemiplegia and the elderly in daily living activities. For this motion, it is difficult to obtain effective features to create a prediction model in environments where large-scale sensor system installation is not feasible and the motion time is short. We performed motion-collection experiments, revealed the features of the target motion and built a prediction model using the multimodal motion features and deep learning. The proposed model achieved an accuracy of 93% macro average and F1-score of 0.69 for a 9-class classification prediction at 35% of the motion completion.

Efficient key distribution for secure and energy-optimized communication in wireless sensor network using bioinspired algorithms

Wireless Sensor Networks (WSNs) play a major part in numerous applications such as smart agriculture, healthcare, and environmental monitoring. Safeguarding protected communication in this network is dominant. Securing data transmission in WSNs needs a strong key distribution device to defend against malicious attacks as well as illegal access. Traditional techniques like pre-shared or centralized key management are often unreasonable owing to resource limitations, particularly in large-scale sensor systems. To overcome this challenge, a lightweight key distribution technique is employed for safeguarding the security and privacy of data transmission streamlining processes decreasing computational overhead as well as energy consumption. By optimizing and simplifying key distribution devices, we propose to improve the complete efficacy and trustworthiness of WSNs that aid safe communication while preserving valuable energy resources. Therefore, this article designs an Efficient Key Distribution for Secure and Energy-Optimized Communication using Bioinspired Algorithms (EKD-SOCBA) for WSN. The purpose of the EKD-SOCBA technique is to accomplish security and energy efficiency in WSNs. Initially, the EKD-SOCBA technique applies a golden jackal optimization (GJO) based clustering approach to cluster the nodes and select cluster heads (CHs). Also, a lightweight Dynamic Step-wise Tiny Encryption Algorithm (DS-TEA) is applied to secure data transmission in the network. Finally, a lightweight key management phase is employed to protect the encryption key and decrease energy utilization and overhead costs. To exhibit the enhanced act of the EKD-SOCBA model, a comprehensive set of imitations was involved. Extensive results stated enhanced presentation of EKD-SOCBA methodology over other models on WSN.

Maintenance planning recommendation of complex industrial equipment based on knowledge graph and graph neural network

Maintenance planning is a significant part of predictive maintenance, which involves task planning, resource scheduling, and prevention. With large-scale sensor systems in modern factories, much data will be captured during monitoring and maintenance of complex industrial equipment. Accumulated data facilitates maintenance planning becomes more thorough and timely. Recently, a knowledge graph (KG) was offered to handle large-scale, unorganized maintenance data semantically, resulting in better data usage. Some prior studies have utilized KG for maintenance planning with semantic searching or graph structure-based algorithms, nevertheless neglecting the prediction of potential linkage. To fill this gap, a maintenance-oriented KG is established firstly based on a well-defined domain-specific ontology schema and accumulated maintenance data. Then, an Attention-Based Compressed Relational Graph Convolutional Network is proposed to predict potential solutions and explain fault in maintenance tasks. Lastly, a maintenance case of oil drilling equipment is carried out, where the proposed model is compared with other cutting-edge models to demonstrate its effectiveness in link prediction. This research is anticipated to shed light on future adoption of KG in maintenance planning recommendations.

Special Issue on the Application of Machine Learning and Artificial Intelligence in Fault Diagnostics

The rapid development of data science and associated artificial intelligence (AI) methods has seen a substantial increase in interest in their application to anomaly detection, fault diagnostic, and prognostic challenges across a wide range of industrial and civil applications. Such approaches may well be the complement that is sought for conventional physical model-based and statistical approaches which often struggle to achieve the desired performance when dealing with complex engineering systems. Researchers have started to apply a range of machine learning and AI-based methods to the large-scale, multi-dimensional data that is often associated with large-scale sensor systems, particularly those which involve IoT devices. There is clear scope for the further development of such approaches, such as deep learning, transfer learning methods, and AI models, to enhance the performance of condition monitoring and associated technologies, and this is the key motivation for this Special Issue. This Special Issue on the Application of Machine Learning and Artificial Intelligence in Fault Diagnostics contains 6 papers.

Providentia – A Large-Scale Sensor System for the Assistance of Autonomous Vehicles and Its Evaluation

The environmental perception of an autonomous vehicle is limited by its physical sensor ranges and algorithmic performance, as well as by occlusions that degrade its understanding of an ongoing traffic situation. This not only poses a significant threat to safety and limits driving speeds, but it can also lead to inconvenient maneuvers. Intelligent Infrastructure Systems can help to alleviate these problems. An Intelligent Infrastructure System can fill in the gaps in a vehicle’s perception and extend its field of view by providing additional detailed information about its surroundings, in the form of a digital model of the current traffic situation, i.e., a digital twin. However, detailed descriptions of such systems and working prototypes demonstrating their feasibility are scarce. In this paper, we propose a hardware and software architecture that enables such a reliable Intelligent Infrastructure System to be built. We have implemented this system in the real world and demonstrate its ability to create an accurate digital twin of an extended highway stretch, thus enhancing an autonomous vehicle’s perception beyond the limits of its onboard sensors. Furthermore, we evaluate the accuracy and reliability of the digital twin by using aerial images and earth observation methods for generating ground truth data.

Multi-Sensor and Decision-Level Fusion-Based Structural Damage Detection Using a One-Dimensional Convolutional Neural Network.

This paper presents a novel approach to substantially improve the detection accuracy of structural damage via a one-dimensional convolutional neural network (1-D CNN) and a decision-level fusion strategy. As structural damage usually induces changes in the dynamic responses of a structure, a CNN can effectively extract structural damage information from the vibration signals and classify them into the corresponding damage categories. However, it is difficult to build a large-scale sensor system in practical engineering; the collected vibration signals are usually non-synchronous and contain incomplete structure information, resulting in some evident errors in the decision stage of the CNN. In this study, the acceleration signals of multiple acquisition points were obtained, and the signals of each acquisition point were used to train a 1-D CNN, and their performances were evaluated by using the corresponding testing samples. Subsequently, the prediction results of all CNNs were fused (decision-level fusion) to obtain the integrated detection results. This method was validated using both numerical and experimental models and compared with a control experiment (data-level fusion) in which all the acceleration signals were used to train a CNN. The results confirmed that: by fusing the prediction results of multiple CNN models, the detection accuracy was significantly improved; for the numerical and experimental models, the detection accuracy was 10% and 16–30%, respectively, higher than that of the control experiment. It was demonstrated that: training a CNN using the acceleration signals of each acquisition point and making its own decision (the CNN output) and then fusing these decisions could effectively improve the accuracy of damage detection of the CNN.

Correction: Large-scale sensor systems based on graphene electrolyte-gated field-effect transistors.

Correction for 'Large-scale sensor systems based on graphene electrolyte-gated field-effect transistors' by Charles Mackin, et al., Analyst, 2016, 141, 2704-2711.

Optimization of Wireless Sensor Networks using PPSS Algorithm

In the development of various large-scale sensor systems, a particularly challenging problem is how to dynamically organize the sensors into a wireless communication network and route sensed information from the field sensors to a target system. The prime motivation of our work is to balance the inherent trade-off between the resource consumption and the accuracy of the target tracking in wireless sensor networks. Toward this objective, the study goes through a new energy-efficient dynamic optimization-based sleep scheduling and target prediction technique for large-scale sensor networks. We present a probability-based prediction and optimization-based sleep scheduling protocol (PPSS) to improve energy efficiency of proactive wake up. A cluster-based scheme is exploited for optimization-based sleep scheduling. At every sampling instant, only one cluster of sensors that located in the proximity of the target is activated, whereas the other sensors are inactive. To activate the most appropriate cluster, we propose a non myopic rule, which is based on not only the target state prediction but also its future tendency. Finally, the effectiveness of the proposed approach is evaluated and compared with the state-of-the-art protocols in terms of tracking accuracy, inter node communication, and computation complexity . Index Terms—Energy efficiency, target prediction, sleep scheduling, target tracking, sensor networks. In a large-scale sensor network, hundreds or thousands of tiny sensor nodes are randomly deployed into a monitoring field to gather data. The complexity of computation and communication increases with the number of active sensor nodes tracking the target. The amount of energy used in the network is proportional to the number of active sensor nodes. It is best for sensor nodes to be arranged into collaborative m groups. Group collaboration should be limited to a tracking area around the target so that the communication and computation will be independent of the size of the network. Multiple nodes surrounding the target may collaborate and gather information. The tracking accuracy and performance is limited to the information in those sensors. In a large-scale sensor network, it is important to locate the target with high accuracy while consuming the least amount of energy. Some of the existing studies have focused on energy efficient methods to track mobile targets. Our objective is to propose a simple routing metric that is composed of the energy expenditure and battery power of a node. Therefore, the cluster activation phase has a great importance not only in minimizing energy consumption but also improve the optimized tracking accuracy.

Engaging older adults in the visualization of sensor data facilitated by an open platform for connected devices.

The use of smart home sensor systems is growing primarily due to the appeal of unobtrusively monitoring older adult health and wellness. However, integrating large-scale sensor systems within residential settings can be challenging when deployment takes place across multiple environments, requiring customization of applications, connection across various devices and effective visualization of complex longitudinal data. The objective of the study was to demonstrate the implementation of a smart home system using an open, extensible platform in a real-world setting and develop an application to visualize data real time. We deployed the open source Lab of Things platform in a house of 11 residents as a demonstration of feasibility over the course of 3 months. The system consisted of Aeon Labs Z-wave Door/Window sensors and an Aeon Labs Multi-sensor that collected data on motion, temperature, luminosity, and humidity. We applied a Rapid Iterative Testing and Evaluation approach towards designing a visualization interface engaging gerontological experts. We then conducted a survey with 19 older adult and caregiver stakeholders to inform further design revisions. Our initial visualization mockups consisted of a bar chart representing activity level over time. Family members felt comfortable using the application. Older adults however, indicated it would be difficult to learn to use the application, and had trouble identifying utility. A key for older adults was ensuring that the data collected could be utilized by their family members, physicians, or caregivers. The approach described in this work is generalizable towards future smart home deployments and can be a valuable guide for researchers to scale a study across multiple homes and connected devices, and to create personalized interfaces for end users.

Large-scale sensor systems based on graphene electrolyte-gated field-effect transistors.

This work reports a novel graphene electrolyte-gated field-effect transistor (EGFET) array architecture along with a compact, self-contained, and inexpensive measurement system that allows DC characterization of hundreds of graphene EGFETs as a function of VDS and VGS within a matter of minutes. We develop a reliable graphene EGFET fabrication process capable of producing 100% yield for a sample size of 256 devices. Large sample size statistical analysis of graphene EGFET electrical performance is performed for the first time. This work develops a compact piecewise DC model for graphene EGFETs that is shown capable of fitting 87% of IDSvs. VGS curves with a mean percent error of 7% or less. The model is used to extract variations in device parameters such as mobility, contact resistance, minimum carrier concentration, and Dirac point. Correlations in variations are presented. Lastly, this work presents a framework for application-specific optimization of large-scale sensor designs based on graphene EGFETs.

R2: Incremental Reprogramming Using Relocatable Code in Networked Embedded Systems

We present R2, an incremental reprogramming approach using relocatable code, to improve program similarity for efficient incremental reprogramming in networked embedded systems. R2 achieves a higher degree of similarity than existing approaches by mitigating effects of both function shifts and data shifts. R2 adopts a content-aware differencing algorithm to generate small delta files for efficient dissemination. Besides, it makes efficient use of memory and does not degrade program quality. We implement R2 based on TinyOS 2.1 and demonstrate its advantages through detailed analysis of TinyOS examples. We also present case studies on the software programs of a large-scale sensor system GreenOrbs. Results show that R2 reduces the dissemination cost by approximately 65 percent compared to state-of-the-art network reprogramming approach Deluge, and reduces the dissemination cost by approximately 20 percent compared to Zephyr and Hermes the latest works on incremental reprogramming.

A dynamic clustering and energy efficient routing technique for sensor networks

In the development of various large-scale sensor systems, a particularly challenging problem is how to dynamically organize the sensors into a wireless communication network and route sensed information from the field sensors to a remote base station. This paper presents a new energy-efficient dynamic clustering technique for large-scale sensor networks. By monitoring the received signal power from its neighboring nodes, each node estimates the number of active nodes in realtime and computes its optimal probability of becoming a cluster head, so that the amount of energy spent in both intra- and inter-cluster communications can be minimized. Based on the clustered architecture, this paper also proposes a simple multihop routing algorithm that is designed to be both energy-efficient and power-aware, so as to prolong the network lifetime. The new clustering and routing algorithms scale well and converge fast for large-scale dynamic sensor networks, as shown by our extensive simulation results.

Long-distance fiber grating sensor system using a fiber ring laser with EDWA and SOA

This investigation presents and demonstrates the feasibility of a novel fiber Bragg grating (FBG) sensor system using a fiber ring laser with a hybrid amplifier. The hybrid amplifier comprises an erbium-doped waveguide amplifier (EDWA) and a semiconductor optical amplifier (SOA). Experimental results demonstrate that the hybrid amplifier has a higher amplifier spontaneous emission power and gain spectrum. Moreover, this novel fiber ring laser provides a stable multiwavelength output with an optical signal-to-noise ratio over 50 dB, even when FBGs are located at a 25 km remote sensing position. The primary feature of the proposed fiber laser is that it facilitates a long-distance or a large-scale sensor system.

Internet-scale sensor systems: Design and policy

As seen in this issue of pervasive computing, sensor networks are the subject of active research. Faculty members at a number of institutions are incorporating research results into new courses on sensor networks and related issues. Carnegie Mellon University has done just that in an innovative graduate course on large-scale sensor systems. Covering a broad set of multidisciplinary topics, this course represents a successful university-industry partnership. The course would pay particular attention to problems associated with data-mining Web cam information, as in the IrisNet project.