Sensor Data Fusion Techniques Research Articles

Utilizing an Arduino Uno-Based System with Integrated Sensor Data Fusion and Filtration Techniques for Enhanced Air Quality Monitoring in Residential Spaces

This study presents an air quality monitoring system that employs the Arduino Uno microcontroller. The system is augmented with a moving average filter and data fusion techniques from BME680 and CCS811 sensors, which are designed to process and combine data from these sensors. The system was tested and analyzed empirically across a range of residential environments in order to validate its efficacy. The findings indicated that the typical IAQ level in a bedroom was approximately 20 units. However, this level increased significantly, reaching 140 units, within minutes after the introduction of a 17% perfume spray. In contrast, the use of an aromatic diffuser resulted in a smaller increase in IAQ to 40 units, which returned to normal levels after ventilation. Moreover, the analysis demonstrated that the kitchen and bathroom exhibited inferior air quality in comparison to the bedroom. This was evidenced by elevated VOC and humidity levels, which were observed to be 10–20% higher due to the combined effects of household activities and inadequate ventilation. This study makes a significant contribution to the field of air quality monitoring by proposing a solution that employs sensor technology and data processing methods to enhance the quality of life within residential spaces.

Integration of blockchain and machine learning for safe and efficient autonomous car systems: A survey

The integration of blockchain and machine learning technologies has the potential to enable the development of more secure, reliable, and efficient autonomous car systems. Blockchain can be used to store, manage, and share the large amounts of data generated by autonomous vehicle various sensors and cameras, ensuring the integrity and security of these data. Machine learning algorithms can be used to analyze and fuse these data in real time, allowing the vehicle to make informed decisions about how to navigate its environment and respond to changing conditions. Thus, the combination of these technologies has the potential to improve the safety, performance, and scalability of autonomous car systems, making them a more applicable and attractive option for consumers and industry stakeholders. In this paper, all relevant technologies, such as machine learning, blockchain and autonomous cars, were explored. Various techniques of machine learning were investigated, including reinforcement learning strategies, the evolution of artificial neural networks and main deep learning algorithms. The main features of the blockchain technology, as well as its different types and consensus mechanisms, were discussed briefly. Autonomous cars, their different types of sensors, potential vulnerabilities, sensor data fusion techniques, and decision-making models were addressed, and main problem domains and trends were underlined. Furthermore, relevant research discussing blockchain for intelligent transportation systems and internet of vehicles was examined. Subsequently, papers related to the integration of blockchain with machine learning for autonomous cars and vehicles were compared and summarized. Finally, the main applications, challenges and future trends of this integration were highlighted.

Condition monitoring of Industrial machine’s transmission elements – MSDF feature extraction approach

ABSTRACT Predictive maintenance is a modern Industry 4.0 strategy in which machines are continuously monitored to detect flaws and prevent breakdowns before they occur. A single sensor focuses on a single variable while neglecting larger information components, resulting in poor data quality and an increased risk of issues in critical rotating equipment. Multi-sensory configuration technology has been developed to collect huge amounts of data from a machine in order to enhance monitoring capabilities in terms of precision, resolution, efficiency, resilience, and trustworthiness of the overall system. The goal of this work is to provide an integrated perspective on machine monitoring using the Multi Sensor Data Fusion (MSDF) technique. On four fault bearings, a case study contrasts the results of single and multiple sensors. A feature-level data fusion method is used, in which computations using time-domain vibration signature data are utilised to build a fusional vector, which is then classified using SVM and analysed with Gaussian kernels. The experimental results suggest that the proposed Gaussian kernel with SVM technique outperforms single sensor data interpretation in terms of classification accuracy and generalisation capability. It is an efficient way for finding defects in rotating machinery in excessively noisy environments.

Sensor fusion applied to the estimate of luminous intensity (LUX) in practical class

In the last ten years, the development of sensors with greater accuracy and precision due to improvements in manufacturing processes has enabled the expansion of their use in several areas. However, the purchase price, mainly of products from renowned manufacturers, in view of their applications, can make simpler projects unfeasible. The sensor data fusion technique is a viable alternative to resolve this issue, as mathematical models can be proposed and used in different situations. These models allow improving the data obtained in order to generate reliable information. Therefore, the objective of this work was to verify the performance of multiple linear regression applied to the fusion of redundant quantitative data from 5mm LDR sensors in estimating the luminous intensity (LUX) in simulated scenarios. To carry out the experiment, 3 LDR (Light Dependent Resistor) sensors, 3 LM393 signal conditioners, 1 USB 6009 DAQ data acquisition board (14 bits), 1 LT40 Extech luxmeter, in addition to the LabView software were used. It was found that LDR A and B sensors showed higher levels of accuracy. Furthermore, a significant improvement in the level of accuracy was found when combining data from sensors A and B in the form of multiple linear regression.

Sensor data fusion techniques in the construction of generalized VORONOI graph for on-line motion planning in robot navigation

Sensor data fusion techniques in the construction of generalized VORONOI graph for on-line motion planning in robot navigation

Three-dimensional wind field reconstruction using tucker decomposition with optimal sensor placement

Short-term wind prediction is of great significance for wind power trading strategy, equipment protection, and effective control. Physical models can provide accurate and detailed wind speed data but require a large amount of time for calculation. To solve this problem, based on Tucker decomposition, a sensor and CFD (Computational Fluid Dynamics) data fusion technique is proposed, which allows us to readily obtain an estimation of the ‘actual’ three-dimensional wind field from sensor observations. Moreover, a new greedy algorithm based on minimum condition is developed to stabilize the inverse process and derive optimal sensor placement. In the simulation, the ‘actual’ wind field can be reconstructed with reasonable accuracy for different wind shear characteristics, and the relative errors of absolute velocity is less than 0.2%. Additionally, the reconstruction time is much less than ultra-short-term forecasting and accounting for only 0.2%–3.2% of the CFD calculation time. The experimental results show that the optimal placement is much more effective than random placement in terms of minimizing the relative error (the relative error can be controlled within 6%). In general, the innovative approach combines the advantages of both statistical and physical approaches while compensating for their shortcomings, making it potentially valuable in short-term wind forecasting.

Detecting Faults at the Edge via Sensor Data Fusion Echo State Networks.

The pervasive use of sensors and actuators in the Industry 4.0 paradigm has changed the way we interact with industrial systems. In such a context, modern frameworks are not only limited to the system telemetry but also include the detection of potentially harmful conditions. However, when the number of signals generated by a system is large, it becomes challenging to properly correlate the information for an effective diagnosis. The combination of Artificial Intelligence and sensor data fusion techniques is a valid solution to address this problem, implementing models capable of extracting information from a set of heterogeneous sources. On the other hand, the constrained resources of Edge devices, where these algorithms are usually executed, pose strict limitations in terms of memory occupation and models complexity. To overcome this problem, in this paper we propose an Echo State Network architecture which exploits sensor data fusion to detect the faults on a scale replica industrial plant. Thanks to its sparse weights structure, Echo State Networks are Recurrent Neural Networks models, which exhibit a low complexity and memory footprint, which makes them suitable to be deployed on an Edge device. Through the analysis of vibration and current signals, the proposed model is able to correctly detect the majority of the faults occurring in the industrial plant. Experimental results demonstrate the feasibility of the proposed approach and present a comparison with other approaches, where we show that our methodology is the best trade-off in terms of precision, recall, F1-score and inference time.

An enhanced energy proficient clustering (EEPC) algorithm for relay selection in heterogeneous WSNs

The recent advancements in Wireless Sensor Networks (WSNs) have brought attention to the field of sensor tracking events. Habitat monitoring is considered as one of the most important applications of WSNs, which ensures wildlife conservation. Researchers have proposed various solutions to select the optimal path in the field of sensor tracking. However, energy dissipation of sensor nodes and fault tolerance during data transformation is still one of the major challenges of the WSN environment in dynamic scenarios. In this paper, an Enhanced Energy Proficient Clustering (EEPC) is proposed to reduce the energy consumption of the entire sensor nodes in the field of tracking events. The network is created with both fixed and mobile nodes. Initially, fixed nodes broadcast information, and mobile nodes select the cluster head from fixed nodes. The mobile nodes select their cluster head (CH) based on their associated placement and energy level. Mobile sensor nodes (SNs) transmit data to the CH. It introduces the concept of finding relay nodes, which are fixed nodes. The EEPC algorithm selects the relay nodes based on their velocity and location by calculating particle fitness value. The selected intermediate relay nodes transmit the collected information to the Base Station (BS) using the sensor data fusion technique. The link fault of nodes could be predicted based on the deviation value. The simulation results show that the proposed approach points out progress with reference to various performance metrics. The proposed work minimizes the energy depletion and enhances the network lifetime compared to other existing protocols.

An Overview of IoT Sensor Data Processing, Fusion, and Analysis Techniques.

In the recent era of the Internet of Things, the dominant role of sensors and the Internet provides a solution to a wide variety of real-life problems. Such applications include smart city, smart healthcare systems, smart building, smart transport and smart environment. However, the real-time IoT sensor data include several challenges, such as a deluge of unclean sensor data and a high resource-consumption cost. As such, this paper addresses how to process IoT sensor data, fusion with other data sources, and analyses to produce knowledgeable insight into hidden data patterns for rapid decision-making. This paper addresses the data processing techniques such as data denoising, data outlier detection, missing data imputation and data aggregation. Further, it elaborates on the necessity of data fusion and various data fusion methods such as direct fusion, associated feature extraction, and identity declaration data fusion. This paper also aims to address data analysis integration with emerging technologies, such as cloud computing, fog computing and edge computing, towards various challenges in IoT sensor network and sensor data analysis. In summary, this paper is the first of its kind to present a complete overview of IoT sensor data processing, fusion and analysis techniques.

A review on multi sensor data fusion technique in CNC machining of tailor-made nanocomposites

In this study, a multiple sensor data fusion system is anticipated as essential for monitoring of cutting operations, by identifying suitable sensor locations to obtain feedback signals periodically. The sensor signal derives the failure during machining owing to complex cutting tool geometry even in machining of composites. Nano metal matrix composites (NMMC) being extremely upright in mechanical characteristics, consequently machining of these hybrid nano metal matrix composites reinforced with difficult-to-cut nano particles leads to reduced tool life thereby causing rapid flank wear. Therefore, it is a challenge to identify the wear features caused during machining of tailor made NMMC’s, reducing wastage and preventing machine malfunction. This paper presents a comprehensive review on the machining strategies in extreme output conditions which rely on input parameters of speed, feed and depth of cut influencing tool life during CNC machining. This can be achieved only with multiple sensor data fusion technique during CNC machining.

Smart Aging System: Uncovering the Hidden Wellness Parameter for Well-Being Monitoring and Anomaly Detection.

Background: Ambiguities and anomalies in the Activity of Daily Living (ADL) patterns indicate deviations from Wellness. The monitoring of lifestyles could facilitate remote physicians or caregivers to give insight into symptoms of the disease and provide health improvement advice to residents; Objective: This research work aims to apply lifestyle monitoring in an ambient assisted living (AAL) system by diagnosing conduct and distinguishing variation from the norm with the slightest conceivable fake alert. In pursuing this aim, the main objective is to fill the knowledge gap of two contextual observations (i.e., day and time) in the frequent behavior modeling for an individual in AAL. Each sensing category has its advantages and restrictions. Only a single type of sensing unit may not manage composite states in practice and lose the activity of daily living. To boost the efficiency of the system, we offer an exceptional sensor data fusion technique through different sensing modalities; Methods: As behaviors may also change according to other contextual observations, including seasonal, weather (or temperature), and social interaction, we propose the design of a novel activity learning model by adding behavioral observations, which we name as the Wellness indices analysis model; Results: The ground-truth data are collected from four elderly houses, including daily activities, with a sample size of three hundred days plus sensor activation. The investigation results validate the success of our method. The new feature set from sensor data fusion enhances the system accuracy to (98.17% ± 0.95) from (80.81% ± 0.68). The performance evaluation parameters of the proposed model for ADL recognition are recorded for the 14 selected activities. These parameters are Sensitivity (0.9852), Specificity (0.9988), Accuracy (0.9974), F1 score (0.9851), False Negative Rate (0.0130).

Harnessing Commodity Wearable Devices to Capture Learner Engagement

With the aim of capturing learner engagement, we propose and advocate the use of commodity wearable devices and their built-in sensors (e.g., accelerometer, gyroscope, and magnetic sensor) to detect the fine-grained learning activities (e.g., writing notes or raising the hand in class). Next, by leveraging the established theory that links learner activities to learner engagement, the detected learner activities can be used to further infer the learner engagement levels, durations, and other key information. We thus designed a hassle-free and non-intrusive system running on the latest wrist-worn commodity wearable devices, which adopts the latest activity recognition and sensor data fusion techniques. We conducted the system-level evaluation, survey, and interviews involving both students and teachers. The evaluation results show that our system can accomplish the accurate learner activity recognition task, and meanwhile effectively capture the learner engagement. We also provide the engagement-based intervention service during class to illustrate the unique usefulness of the proposed system.

Hand Gesture Recognition in Automotive Human⁻Machine Interaction Using Depth Cameras.

In this review, we describe current Machine Learning approaches to hand gesture recognition with depth data from time-of-flight sensors. In particular, we summarise the achievements on a line of research at the Computational Neuroscience laboratory at the Ruhr West University of Applied Sciences. Relating our results to the work of others in this field, we confirm that Convolutional Neural Networks and Long Short-Term Memory yield most reliable results. We investigated several sensor data fusion techniques in a deep learning framework and performed user studies to evaluate our system in practice. During our course of research, we gathered and published our data in a novel benchmark dataset (REHAP), containing over a million unique three-dimensional hand posture samples.

Development of a High Precision Displacement Measurement System by Fusing a Low Cost RTK-GPS Sensor and a Force Feedback Accelerometer for Infrastructure Monitoring.

A displacement measurement system fusing a low cost real-time kinematic global positioning system (RTK-GPS) receiver and a force feedback accelerometer is proposed for infrastructure monitoring. The proposed system is composed of a sensor module, a base module and a computation module. The sensor module consists of a RTK-GPS rover and a force feedback accelerometer, and is installed on a target structure like conventional RTK-GPS sensors. The base module is placed on a rigid ground away from the target structure similar to conventional RTK-GPS bases, and transmits observation messages to the sensor module. Then, the initial acceleration, velocity and displacement responses measured by the sensor module are transmitted to the computation module located at a central monitoring facility. Finally, high precision and high sampling rate displacement, velocity, and acceleration are estimated by fusing the acceleration from the accelerometer, the velocity from the GPS rover, and the displacement from RTK-GPS. Note that the proposed displacement measurement system can measure 3-axis acceleration, velocity as well as displacement in real time. In terms of displacement, the proposed measurement system can estimate dynamic and pseudo-static displacement with a root-mean-square error of 2 mm and a sampling rate of up to 100 Hz. The performance of the proposed system is validated under sinusoidal, random and steady-state vibrations. Field tests were performed on the Yeongjong Grand Bridge and Yi Sun-sin Bridge in Korea, and the Xihoumen Bridge in China to compare the performance of the proposed system with a commercial RTK-GPS sensor and other data fusion techniques.

Development of a tactile sensor based on optical fiber specklegram analysis and sensor data fusion technique

The development of a tactile sensor based on optical fiber specklegram analysis is reported. The device is comprised of 9 microbending transducers connected to 3 multimode fibers, and attached to a 30×30mm2 touching surface in a matrix arrange. The output fiber speckle fields, produced by coherent light transmission through the multimode waveguides, are processed for evaluation of the normalized inner-product coefficients, being further correlated to the external load characteristics according to a specklegram referencing approach. Finally, the magnitude and location of the forces applied over the tactile frame are estimated by means of a data fusion technique, yielding the probabilistic distribution of the mechanical disturbances. The sensor characterization indicated a 0.5N−1 sensitivity, with capability to detect variations of 1mm in force location, being comparable to the requirements for tactile technologies. Moreover, the data fusion approach provided the estimation of the load distribution with 0.86 relative accuracy, allowing the assessment of tactile variables based on the response of only 3 sensing fibers and without utilizing wavelength multiplexing schemes.

A Systematic Approach of Data Fusion Technique in RFID Sensor Network using Neuro-Fuzzy Technique

In this paper, a systematic approach is used for sale prediction in a multistoried retail business with the help of multi sensor data fusion technique using Neural Network and Fuzzy Logic. This method can better solve problems existing in traditional sale prediction which are basically depends on the personal experience. In this work a 3-layers data fusion structure is used. In this system, the sale data experiential characteristic and the sale data-fitting characteristic are fused by fuzzy inference system to get sale prediction. After using the Feed Forward Back Propagation algorithm, the system is trained for predefined target value and then the system calculate the sale statistic in runtime which is fused with the data of expert databases using fuzzy logic technique.

From Data Acquisition to Data Fusion: A Comprehensive Review and a Roadmap for the Identification of Activities of Daily Living Using Mobile Devices.

This paper focuses on the research on the state of the art for sensor fusion techniques, applied to the sensors embedded in mobile devices, as a means to help identify the mobile device user’s daily activities. Sensor data fusion techniques are used to consolidate the data collected from several sensors, increasing the reliability of the algorithms for the identification of the different activities. However, mobile devices have several constraints, e.g., low memory, low battery life and low processing power, and some data fusion techniques are not suited to this scenario. The main purpose of this paper is to present an overview of the state of the art to identify examples of sensor data fusion techniques that can be applied to the sensors available in mobile devices aiming to identify activities of daily living (ADLs).

Quantitative characterization of pulverized coal and biomass–coal blends in pneumatic conveying pipelines using electrostatic sensor arrays and data fusion techniques

Quantitative data about the dynamic behaviour of pulverized coal and biomass–coal blends in fuel injection pipelines allow power plant operators to detect variations in fuel supply and oscillations in the flow at an early stage, enable them to balance fuel distribution between fuel feeding pipes and ultimately to achieve higher combustion efficiency and lower greenhouse gas emissions. Electrostatic sensor arrays and data fusion algorithms are combined to provide a non-intrusive solution to the measurement of fuel particle velocity, relative solid concentration and flow stability under pneumatic conveying conditions. Electrostatic sensor arrays with circular and arc-shaped electrodes are integrated in the same sensing head to measure ‘averaged’ and ‘localized’ characteristics of pulverized fuel flow. Data fusion techniques are applied to optimize and integrate the results from the sensor arrays. Experimental tests were conducted on the horizontal section of a 150 mm bore pneumatic conveyor circulating pulverized coal and sawdust under various flow conditions. Test results suggest that pure coal particles travel faster and carry more electrostatic charge than biomass–coal blends. As more biomass particles are added to the flow, the overall velocity of the flow reduces, the electrostatic charge level on particles decreases and the flow becomes less stable compared to the pure coal flow.

Fuzzy Logic Based Sensors Data Fusion for Mobile Robot Navigation

In this paper, a sensor data fusion technique is proposed to develop the effective robot navigation and optimize the navigation rules using a Modified Fuzzy Associative Memory (MFAM). MFAM provides good flexibility to use multiple input space and reduction of rule base for robot navigation. The behavior rules obtained from MFAM model are tested using the experimental studies and results are discussed and compared with the existing methods.

Simulation study of artificial ocular movement with intelligent control

This paper is concerned with the designing and construction of a biomedical assistive device for ophthalmic patients to regain the natural movement of their ocular prostheses. Various aspects of the device are discussed in this paper in terms of sensor design, sensor data fusion techniques, artificial ocular actuation, control system and experimental prototype setup. Simulation and experimental study are included to illustrate the effectiveness of the device. The paper concludes with future considerations.