Non-Contact Optical Detection of Foreign Materials Adhered to Color Filter and Thin-Film Transistor

This paper describes the non-contact optical detection of debris material that adheres to the substrates of color filters (CFs) and thin-film transistors (TFTs) by area charge-coupled devices (CCDs) and laser sensors. One of the optical detections is a side-view illumination by an area CCD that emits a coherency light to detect debris on the CF. In contrast to the height of the debris material, the image is acquired by transforming the geometric shape from a square to a circle. As a result, the side-view illumination from the area CCD identified the height of the debris adhered to the black matrix (BM) as well as the red, green, and blue of a CF with 95, 97, 98, and 99% accuracy compared to the golden sample. The uncertainty analysis was at 5% for the BM, 3% for the red, 2% for the green, and 1% for the blue. The other optical detection, a laser optical interception with a horizontal alignment, inspected the material foreign to the TFT. At the same time, laser sensors intercepted the debris on the TFT at a voltage of 3.5 V, which the five sets of laser optics make scanning the sample. Consequently, the scanning rate reached over 98% accuracy, and the uncertainty analysis was within 5%. Thus, both non-contact optical methods can detect debris at a 50 μm height or lower. The experiment presents a successful design for the efficient prevention of a valuable component malfunction.

Pallet Recognition with Multi-Task Learning for Automated Guided Vehicles

As the need for efficient warehouse logistics has increased in manufacturing systems, the use of automated guided vehicles (AGVs) has also increased to reduce travel time. The AGVs are controlled by a system using laser sensors or floor-embedded wires to transport pallets and their loads. Because such control systems have only predefined palletizing strategies, AGVs may fail to engage incorrectly positioned pallets. In this study, we consider a vision sensor-based method to address this shortcoming by recognizing a pallet’s position. We propose a multi-task deep learning architecture that simultaneously predicts distances and rotation based on images obtained from a visionary sensor. These predictions complement each other in learning, allowing a multi-task model to learn and execute tasks impossible with single-task models. The proposed model can accurately predict the rotation and displacement of the pallets to derive information necessary for the control system. This information can be used to optimize a palletizing strategy. The superiority of the proposed model was verified by an experiment on images of stored pallets that were collected from a visionary sensor attached to an AGV.

Analysis of Laser Sensors and Camera Vision in the Shoe Position Inspection System

Inspection systems are currently an evolving field in the industry. The main goal is to provide a picture of the quality of intermediates and products in the production process. The most widespread sensory system is camera equipment. This article describes the implementation of camera devices for checking the location of the upper on the shoe last. The next part of the article deals with the analysis of the application of laser sensors in this task. The results point to the clear advantages of laser sensors in the inspection task of placing the uppers on the shoe’s last. The proposed method defined the resolution of laser scanners according to the type of scanned surface, where the resolution of point cloud ranged from 0.16 to 0.5 mm per point based on equations representing specific points approximated to polynomial regression in specific places, which are defined in this article. Next, two inspection systems were described, where one included further development in the field of automation and industry 4.0 and with a high perspective of development into the future. The main aim of this work is to conduct analyses of sensory systems for inspection systems and their possibilities for further work mainly based on the resolution and quality of obtained data. For instance, dependency on scanning complex surfaces and the achieved resolution of scanned surfaces.

Development of a novel tool for automation of the contamination measurement

Considering that about 100 000 m2 of wall area per nuclear facility must be decontaminated (Hübner et al., 2017), the automation of mechanical decontamination work offers high potential to support people in performing their work and reduce errors in the decommissioning process. Furthermore, the exposure potential for people in contaminated environments is reduced and they are protected from health hazards (Petereit et al., 2019). In the ROBDEKON project, a competence center is being established in Germany to develop practical robotic systems for decontamination work in hazardous environments. To this end, four research institutions are working with industrial partners on the development of (partially) autonomous robotic systems for the decommissioning and decontamination of nuclear facilities, the handling of waste, and the remediation of landfills and contaminated sites (Petereit et al., 2019). At the Institute for Technology and Management in Construction (KIT-TMB), the focus is on development of an automated solution for the (clearance) measurement of near-surface contaminations. A mobile elevating working platform is used as the robotic platform with a contamination array as the tool. The array measures the surface activity on the wall areas and verifies compliance with the thresholds. The contamination array is based on two sensor concepts: measurement and localization. Up to four hand-held contamination-measuring devices are attached to the array to parallelize the measurement. In order to avoid damaging the sensitive detector window foil of the contamination probes, the wall surface to be measured is first examined for imperfections with the help of a laser scanner. For localization of the array, up to four laser sensors are used for distance measurements. Measurement results are automatically saved after each measurement in a table specific to the measurement method and are available to users for documentation purposes at any time. In the further course of the project, the measurement results depending on the radiation level will be overlaid with a geometric 3D environment mapping.

Joint Soft–Hard Attention for Self-Supervised Monocular Depth Estimation

In recent years, self-supervised monocular depth estimation has gained popularity among researchers because it uses only a single camera at a much lower cost than the direct use of laser sensors to acquire depth. Although monocular self-supervised methods can obtain dense depths, the estimation accuracy needs to be further improved for better applications in scenarios such as autonomous driving and robot perception. In this paper, we innovatively combine soft attention and hard attention with two new ideas to improve self-supervised monocular depth estimation: (1) a soft attention module and (2) a hard attention strategy. We integrate the soft attention module in the model architecture to enhance feature extraction in both spatial and channel dimensions, adding only a small number of parameters. Unlike traditional fusion approaches, we use the hard attention strategy to enhance the fusion of generated multi-scale depth predictions. Further experiments demonstrate that our method can achieve the best self-supervised performance both on the standard KITTI benchmark and the Make3D dataset.

High-Occupancy Vehicle Lane Enforcement System

Introduction: An automatic High-Occupancy Vehicle (HOV) lane enforcement system is developed and evaluated. Current manual enforcement practices by the police bring about safety concerns and unnecessary traffic delays. Only vehicles with more than five passengers are permitted to use HOV lanes on freeways in Korea. Hence, detecting the number of passengers in HOVs is a core element for their development. Methods: For a quick detection capability, a YOLO-based passenger detection model was built. The system comprises three infrared cameras: two are for compartment detection and the other is for number plate recognition. Multiple infrared illuminations with the same frequency as the cameras and laser sensors for vehicle detection and speed measurement are also employed. Results: The performance of the developed system is evaluated with real-world data collected on proving ground. As a result, it showed a passenger detection error of nine percent on average. The performances revealed no difference in vehicle speeds and the number of passengers according to ANOVA tests. Conclusion: Using the developed system, more efficient and safer HOV lane enforcement practices can be made.

RANCANG BANGUN PROTOTYPE SISTEM PENDETEKSI PELANGGARAN PADA ZEBRA CROSS DI LAMPU LALU LINTAS BERBASIS ARDUINO

Current road users cannot be separated from the number of violators, therefore traffic lights are made to regulate traffic on the road. At traffic lights, there is also a zebra crossing which serves as a means of crossing the road for pedestrians. To minimize violations at road intersections, researchers designed a tool to detect traffic violations. Traffic violation detection tool is made in prototype form using a control system with Arduino nano and software. This traffic light system uses LDR and laser sensors to detect these violations by cutting the laser which sends light to the LDR. This tool is also equipped with a webcam camera that functions to photograph violations that occur and a buzzer that functions as a warning to officers and riders in the event of a violation with an average response speed of the webcam of 2.37 seconds and the average response speed of the buzzer is 0.4 seconds. . The snapshot from the webcam is saved automatically on your PC / Laptop.

Experimental Characterization of Lithium-Ion Cell Strain Using Laser Sensors

The characterization of thickness change during operation of LFP/Graphite prismatic batteries is presented in this work. In this regard, current rate dependence, hysteresis behaviour between charge and discharge and correlation with phase changes are deepened. Experimental tests are carried out with a battery testing equipment correlated with optical laser sensors to evaluate swelling. Furthermore, thickness change is computed analytically with a mathematical model based on lattice parameters of the crystal structures of active materials. The results of the model are validated with experimental data. Thickness change is able to capture variations of the internal structure of the battery, referred to as phase change, characteristic of a certain state of charge. Furthermore, phase change shift is a characteristic of battery ageing. Being able to capture these properties with sensors mounted on the external surface the cell is a key feature for improving state of charge and state of health estimation in battery management system.

Laser Fluorescence and Extinction Methods for Measuring the Flow and Composition of Milk in a Milking Machine

Automation of milking systems is linked to accurate measurement of fluctuations in milk flow during milking. To assess the fluctuations of the milk flow, the formation and movement of milk portions in the milking machine-milk pipeline system was studied. By considering the movement of a milk plug along the milk pipeline, a hydraulic model of the formation of a critical volume of milk in the milking machine manifold was compiled. In practice, the most expedient way of determining milk flow parameters may be to measure the laser fluorescent and extinction responses of moving air-milk mixture. We have implemented a new laser sensing method for measuring the flow rate and composition of milk on the basis of counting the optical response pulses received from moving dispersed components by a CCD array or a randomized fiber optic bundle. Using the developed laser sensors, the theoretical model of milk flow was tested.

Integrating Pavement Sensing Data for Pavement Condition Evaluation

Highway pavements are usually monitored in terms of their surface performance assessment, since the major cause that triggers maintenance is reduced pavement serviceability due to surface distresses, excessive pavement unevenness and/or texture loss. A common way to detect pavement surface condition is by the use of vehicle-mounted laser sensors that can rapidly scan huge roadway networks at traffic speeds without the need for traffic interventions. However, excessive roughness might sometimes indicate structural issues within one or more pavement layers or even issues within the pavement foundation support. The stand-alone use of laser profilers cannot provide the related agencies with information on what leads to roughness issues. Contrariwise, the integration of multiple non-destructive data leads to a more representative assessment of pavement condition and enables a more rational pavement management and decision-making. This research deals with an integration approach that primarily combines pavement sensing profile and deflectometric data and further evaluates indications of increased pavement roughness. In particular, data including Falling Weight Deflectometer (FWD) and Road Surface Profiler (RSP) measurements are used in conjunction with additional geophysical inspection data from Ground Penetrating Radar (GPR). Based on pavement response modelling, a promising potential is shown that could proactively assist the related agencies in the framework of transport infrastructure health monitoring.