A Machine Vision-Based Pipe Leakage Detection System for Automated Power Plant Maintenance.

Industrial pipework maintenance inspection can be automated through machine vision-based effusion monitoring. However, colorless effusions such as water can be difficult to detect in a complex industrial environment due to weak illumination and poor visibility of the background. This paper deploys the reflective characteristics of effusion and its lower temperature compared to the environment in order to develop an automatic inspection system for power plant pipeworks' maintenance. Such a system is aimed at detecting the colorless fluid effusion based on dual source images and a contour features algorithm. In this respect, a visible light source unit highlights the reflective features of the effusion edge. Meanwhile, high-definition images of the potential effusion are acquired under both visible and infrared lights. A customized image processing procedure extracts the potential effusion features from the infrared image to retrieve the region of interest for segmentation purposes and transfer such information to the visible light image to determine the effusion contour. Finally, a decision-making support tool based on the image contour closure is enabled for classification purposes. The implementation of the proposed system is tested on a real industrial environment. Experimental results show a classification accuracy up to 99%, demonstrating excellent suitability in meeting industrial requirements.

Intelligent Industrial Cleaning: A Multi-Sensor Approach Utilising Machine Learning-Based Regression.

Effectively cleaning equipment is essential for the safe production of food but requires a significant amount of time and resources such as water, energy, and chemicals. To optimize the cleaning of food production equipment, there is the need for innovative technologies to monitor the removal of fouling from equipment surfaces. In this work, optical and ultrasonic sensors are used to monitor the fouling removal of food materials with different physicochemical properties from a benchtop rig. Tailored signal and image processing procedures are developed to monitor the cleaning process, and a neural network regression model is developed to predict the amount of fouling remaining on the surface. The results show that the three dissimilar food fouling materials investigated were removed from the test section via different cleaning mechanisms, and the neural network models were able to predict the area and volume of fouling present during cleaning with accuracies as high as 98% and 97%, respectively. This work demonstrates that sensors and machine learning methods can be effectively combined to monitor cleaning processes.

A NDIR Mid-Infrared Methane Sensor with a Compact Pentahedron Gas-Cell.

In order to improve the performance of the large divergence angle mid-infrared source in gas sensing, this paper aims at developing a methane (CH4) sensor with non-dispersive infrared (NDIR) technology using a compact pentahedron gas-cell. A paraboloid concentrator, two biconvex lenses and five planar mirrors were used to set up the pentahedron structure. The gas cell is endowed with a 170 mm optical path length with a volume of 19.8 mL. The mathematical model of the cross-section and the three-dimension spiral structure of the pentahedron gas-cell were established. The gas-cell was integrated with a mid-infrared light source and a detector as the optical part of the sensor. Concerning the electrical part, a STM32F429 was employed as a microcontroller to generate the driving signal for the IR source, and the signal from the detector was sampled by an analog-to-digital converter. A static volumetric method was employed for the experimental setup, and 20 different concentration CH4 samples were prepared to study the sensor's evaluation, which revealed a 1σ detection limit of 2.96 parts-per-million (ppm) with a 43 s averaging time.

A Multi-Sensor System for Silkworm Cocoon Gender Classification via Image Processing and Support Vector Machine.

Sericulture is traditionally a labor-intensive rural-based industry. In modern contexts, the development of process automation faces new challenges related to quality and efficiency. During the silkworm farming life cycle, a common issue is represented by the gender classification of the cocoons. Improper cocoon separation negatively affects quantity and quality of the yield resulting in disruptive bottlenecks for the productivity. To tackle this issue, this paper proposes a multi sensor system for silkworm cocoons gender classification and separation. Utilizing a load sensor and a digital camera, the system acquires weight and digital images from individual silkworm cocoons. An image processing procedure is then applied to extract significant shape-related features from each image instance, which, combined with the weight data, are provided as inputs to train a Support Vector Machine-based pattern classifier for gender classification. Subsequently, an air blower mechanism and a conveyor system sort the cocoons into their respective bins. The developed system was trained and tested on two different types of silkworm cocoons breeds, respectively CSR2 and Pure Mysore. The system performances are finally discussed in terms of accuracy, robustness and computation time.

Enhanced Clean-In-Place Monitoring Using Ultraviolet Induced Fluorescence and Neural Networks.

Clean-in-place (CIP) processes are extensively used to clean industrial equipment without the need for disassembly. In food manufacturing, cleaning can account for up to 70% of water use and is also a heavy user of energy and chemicals. Due to a current lack of real-time in-process monitoring, the non-optimal control of the cleaning process parameters and durations result in excessive resource consumption and periods of non-productivity. In this paper, an optical monitoring system is designed and realized to assess the amount of fouling material remaining in process tanks, and to predict the required cleaning time. An experimental campaign of CIP tests was carried out utilizing white chocolate as fouling medium. During the experiments, an image acquisition system endowed with a digital camera and ultraviolet light source was employed to collect digital images from the process tank. Diverse image segmentation techniques were considered to develop an image processing procedure with the aim of assessing the area of surface fouling and the fouling volume throughout the cleaning process. An intelligent decision-making support system utilizing nonlinear autoregressive models with exogenous inputs (NARX) Neural Network was configured, trained and tested to predict the cleaning time based on the image processing results. Results are discussed in terms of prediction accuracy and a comparative study on computation time against different image resolutions is reported. The potential benefits of the system for resource and time efficiency in food manufacturing are highlighted.

Topology Optimization of Spatially Compliant Mechanisms with an Isomorphic Matrix of a 3-UPC Type Parallel Prototype Manipulator.

A novel topology optimization approach is proposed in this paper for the design of three rotational degree-of-freedom (DOF) spatially compliant mechanisms, combining the Jacobian isomorphic mapping matrix with the solid isotropic material with penalization (SIMP) topological method. In this approach, the isomorphic Jacobian matrix of a 3-UPC (U: universal joint, P: prismatic joint, C: cylindrical joint) type parallel prototype manipulator is formulated. Subsequently, the orthogonal triangular decomposition and differential kinematic method is applied to uncouple the Jacobian matrix to construct a constraint for topology optimization. Firstly, with respect to the 3-UPC type parallel prototype manipulator, the Jacobian matrix is derived to map the inputs and outputs to be used for initializing the topology optimization process. Secondly, the orthogonal triangular decomposition with the differential kinematic method is used to reconstruct the uncoupled mapping matrix to derive the 3-UPC type parallel prototype manipulator. Finally, a combination of the solid isotropic material with penalization (SIMP) method and the isomorphic mapping matrix is applied to construct the topological model. A typical three rotational DOF spatially compliant mechanism is reported as a numerical example to demonstrate the effectiveness of the proposed method.

Fabrication, Experiments, and Analysis of an LBM Additive-Manufactured Flexure Parallel Mechanism.

Additive manufacturing technology has advantages for realizing complex monolithic structures, providing huge potential for developing advanced flexure mechanisms for precision manipulation. However, the characteristics of flexure hinges fabricated by laser beam melting (LBM) additive manufacturing (AM) are currently little known. In this paper, the fabrication and characterization of a flexure parallel mechanism through the LBM process are reported for the first time to demonstrate the development of this technique. The geometrical accuracy of the additive-manufactured flexure mechanism was evaluated by three-dimensional scanning. The stiffness characteristics of the flexure mechanism were investigated through finite element analysis and experimental tests. The effective hinge thickness was determined based on the parameters study of the flexure parallel mechanism. The presented results highlight the promising outlook of LBM flexure parts for developing novel nanomanipulation platforms, while additional attention is required for material properties and manufacturing errors.

A Novel Algorithm for Determining the Contextual Characteristics of Movement Behaviors by Combining Accelerometer Features and Wireless Beacons: Development and Implementation.

BACKGROUND: Unfortunately, global efforts to promote "how much" physical activity people should be undertaking have been largely unsuccessful. Given the difficulty of achieving a sustained lifestyle behavior change, many scientists are reexamining their approaches. One such approach is to focus on understanding the context of the lifestyle behavior (ie, where, when, and with whom) with a view to identifying promising intervention targets. OBJECTIVE: The aim of this study was to develop and implement an innovative algorithm to determine "where" physical activity occurs using proximity sensors coupled with a widely used physical activity monitor. METHODS: A total of 19 Bluetooth beacons were placed in fixed locations within a multilevel, mixed-use building. In addition, 4 receiver-mode sensors were fitted to the wrists of a roving technician who moved throughout the building. The experiment was divided into 4 trials with different walking speeds and dwelling times. The data were analyzed using an original and innovative algorithm based on graph generation and Bayesian filters. RESULTS: Linear regression models revealed significant correlations between beacon-derived location and ground-truth tracking time, with intraclass correlations suggesting a high goodness of fit (R2=.9780). The algorithm reliably predicted indoor location, and the robustness of the algorithm improved with a longer dwelling time (>100 s; error <10%, R2=.9775). Increased error was observed for transitions between areas due to the device sampling rate, currently limited to 0.1 Hz by the manufacturer. CONCLUSIONS: This study shows that our algorithm can accurately predict the location of an individual within an indoor environment. This novel implementation of "context sensing" will facilitate a wealth of new research questions on promoting healthy behavior change, the optimization of patient care, and efficient health care planning (eg, patient-clinician flow, patient-clinician interaction).