Xiongfeng Yi

A timely inspection of pipelines is of great importance to prevent oil and gas from leaking. This paper develops a novel robotic system that can automatically grab a flange of pipelines and perform bolt looseness inspection autonomously. The robotic system consists of a 4-DOF robotic arm with a stereo camera system on a mobile platform and a pair of piezoceramic transducers (PZTs) mounted on two fingers of the robotic arm. The PZTs act as a smart touch sensor for bolt looseness inspection when… Show more

A timely inspection of pipelines is of great importance to prevent oil and gas from leaking. This paper develops a novel robotic system that can automatically grab a flange of pipelines and perform bolt looseness inspection autonomously. The robotic system consists of a 4-DOF robotic arm with a stereo camera system on a mobile platform and a pair of piezoceramic transducers (PZTs) mounted on two fingers of the robotic arm. The PZTs act as a smart touch sensor for bolt looseness inspection when the fingers grab a flange. An onboard visual tracking algorithm is developed to provide the 3D position of the flange. The flange is first recognized and then localized in a 2D image by YOLO region-based convolutional neural networks. The 2D position in the image is then transferred to the 3D position through a set of calibrated stereo camera parameters. The robotic arm receives proportional gain control signals based on the distance between the finger center and the flange. In the experiment, we install four cameras, apply a feature-based location method, and calculate the average value of history positions to reduce the detection error. When the robotic arm grabs the flange, a stress wave signal is generated by one of the PZTs mounted on the fingers, and the signal received from the other PZT indicates if the flange is tightly bolted or not. Experimental results have validated this autonomous robotic inspection approach. Show less

The kidney biopsy based diagnosis of Lupus Nephritis (LN) is characterized by low inter-observer agreement, with misdiagnosis being associated with increased patient morbidity and mortality. Although various Computer Aided Diagnosis (CAD) systems have been developed for other nephrohistopathological applications, little has been done to accurately classify kidneys based on their kidney level Lupus Glomerulonephritis (LGN) scores. The successful implementation of CAD systems has also been… Show more

The kidney biopsy based diagnosis of Lupus Nephritis (LN) is characterized by low inter-observer agreement, with misdiagnosis being associated with increased patient morbidity and mortality. Although various Computer Aided Diagnosis (CAD) systems have been developed for other nephrohistopathological applications, little has been done to accurately classify kidneys based on their kidney level Lupus Glomerulonephritis (LGN) scores. The successful implementation of CAD systems has also been hindered by the diagnosing physician's perceived classifier strengths and weaknesses, which has been shown to have a negative effect on patient outcomes. We propose an Uncertainty-Guided Bayesian Classification (UGBC) scheme that is designed to accurately classify control, class I/II, and class III/IV LGN (3 class) at both the glomerular-level classification task (26,634 segmented glomerulus images) and the kidney-level classification task (87 MRL/lpr mouse kidney sections). Data annotation was performed using a high throughput, bulk labeling scheme that is designed to take advantage of Deep Neural Network's (or DNNs) resistance to label noise. Our augmented UGBC scheme achieved a 94.5% weighted glomerular-level accuracy while achieving a weighted kidney-level accuracy of 96.6%, improving upon the standard Convolutional Neural Network (CNN) architecture by 11.8% and 3.5% respectively. Show less

The viscoelastic behavior of polymer solutions is commonly measured using oscillating shear rheometry, however, the accuracy of such methods is limited by the oscillating frequency of the equipment and since the relaxation time of the dilute polymer solutions is short, this requires measurement at very high frequencies. Microrheology has been proposed to overcome this technical challenge. Yet the equipment for resolving the statistics of particle displacements in microrheology is expensive. In… Show more

The viscoelastic behavior of polymer solutions is commonly measured using oscillating shear rheometry, however, the accuracy of such methods is limited by the oscillating frequency of the equipment and since the relaxation time of the dilute polymer solutions is short, this requires measurement at very high frequencies. Microrheology has been proposed to overcome this technical challenge. Yet the equipment for resolving the statistics of particle displacements in microrheology is expensive. In this work, we measured the viscoelastic behavior of Methocel solutions at various concentrations using a conventional epi-fluorescence microscope coupled to a high-speed intensified camera. Statistical Particle Tracking is used in analyzing the mean-squared displacement of the dispersive particles. Relaxation times ranging from 0.76 - 9.00 ms and viscoelastic moduli, G' between 11.34 and 3.39 are reported for Methocel solutions of concentrations between 0.063 - 0.5% Show less

Buoyancy control device (BCD) is an essential component for the underwater vehicle to achieve its vertical maneuvering. Previously, a proportional-integral-derivative (PID) controller was developed for a buoyancy device enabled by ionic polymer-metal composite (IPMC) water electrolyzer. With the help of a fluid pressure sensor, the buoyancy device can be stabilized at a specific depth. However, due to the slow gas generation rate, a transition period had to be added into the depth control in… Show more

Buoyancy control device (BCD) is an essential component for the underwater vehicle to achieve its vertical maneuvering. Previously, a proportional-integral-derivative (PID) controller was developed for a buoyancy device enabled by ionic polymer-metal composite (IPMC) water electrolyzer. With the help of a fluid pressure sensor, the buoyancy device can be stabilized at a specific depth. However, due to the slow gas generation rate, a transition period had to be added into the depth control in order to avoid saturating the control input to the IPMC electrolyzer. In this paper, to systematically obtain the transition period, an optimal trajectory planning is developed for the BCD with consideration of slow gas generation rate. An optimal trajectory is obtained by solving a bang-off-bang minimum time optimal control problem with the control input and velocity constraints. Then a state feedback tracking control is developed to track the optimal trajectory. Simulation results have shown that the BCD is stable during the tracking and the control input and velocity are always bounded within the allowable ranges. Show less

This paper develops and demonstrates cooperative collision avoidance control on two robotic fish propelled by a servo motor and an ionic polymer-metal composite (IPMC)-driven fish tail. First, experiments conducted on a servo motor/IPMC-driven fish demonstrate an impulsive turning behavior in the fish’s trajectory under the application of a specific frequency, amplitude of the servo motor, and a constant voltage on the IPMC joint. These experiments validate the ‘back relaxation’ of the IPMC… Show more

This paper develops and demonstrates cooperative collision avoidance control on two robotic fish propelled by a servo motor and an ionic polymer-metal composite (IPMC)-driven fish tail. First, experiments conducted on a servo motor/IPMC-driven fish demonstrate an impulsive turning behavior in the fish’s trajectory under the application of a specific frequency, amplitude of the servo motor, and a constant voltage on the IPMC joint. These experiments validate the ‘back relaxation’ of the IPMC joint by observing the angular velocity and the centripetal acceleration of the fish. This impulsive turning speed due to the ‘back relaxation’ of IPMC joint is subsequently modeled by a transfer function and this transfer function is then integrated into the development of the collision avoidance laws for the fish. The collision avoidance control law utilizes the impulsive turning capability of the robotic fish. An experimental validation of the collision avoidance law is performed. Show less

This paper introduces a robust visual tracking of objects in complex environments with blocking obstacles and light reflection noises. This visual tracking method utilizes a transfer matrix to project image pixels back to real-world coordinates. During the image process, a color and shape test is used to recognize the object and a vector is used to represent the object, which contains the information of orientation and body length of the object. If the object is partially blocked by the… Show more

This paper introduces a robust visual tracking of objects in complex environments with blocking obstacles and light reflection noises. This visual tracking method utilizes a transfer matrix to project image pixels back to real-world coordinates. During the image process, a color and shape test is used to recognize the object and a vector is used to represent the object, which contains the information of orientation and body length of the object. If the object is partially blocked by the obstacles or the reflection from the water surface, the vector predicts the position of the object. During the real-time tracking, a Kalman filter is used to optimize the result. To validate the method, the visual tracking algorithm was tested by tracking a submarine and a fish on the water surface of a water tank, above which three pieces of blur glass were blocking obstacles between the camera and the object. By using this method, the interference from the reflection of the side glass and the fluctuation of the water surface can be also avoided. Show less

This brief addresses the problem of collision avoidance by robotic fish that have a single caudal fin. The fish is hybrid-tailed in that the caudal fin is driven by a double-jointed mechanism. One joint is driven by a servomotor and the second joint is driven by an ionic polymer-metal composite (IPMC) actuator, which is often called an artificial muscle. For this type of robotic fish, collision avoidance is performed by employing a collision cone approach. Within the framework of the collision… Show more

This brief addresses the problem of collision avoidance by robotic fish that have a single caudal fin. The fish is hybrid-tailed in that the caudal fin is driven by a double-jointed mechanism. One joint is driven by a servomotor and the second joint is driven by an ionic polymer-metal composite (IPMC) actuator, which is often called an artificial muscle. For this type of robotic fish, collision avoidance is performed by employing a collision cone approach. Within the framework of the collision cone approach, Lyapunov-based methods are employed to determine analytical expressions of nonlinear guidance laws with which cooperative collision avoidance can be achieved. It is shown how these cooperative collision avoidance laws can be made optimal in the sense of minimizing an energy-like performance index. The conditions under which the developed guidance laws are robust to sensor measurement errors are determined. Simulations and experiments are performed to validate the guidance laws. Show less

Image-based tracking has been widely used to obtain the position and velocity information of a moving object in a 2-dimensional or 3-dimensional space. However, the tracking process is always affected by reflection noises and blocking obstacles in the environment. This paper provides a robust and optimal algorithm for tracking a moving object on the surface of water. First, we create a matrix to project the image pixels back to the real world coordinate. Second, color and shape tests are used… Show more

Image-based tracking has been widely used to obtain the position and velocity information of a moving object in a 2-dimensional or 3-dimensional space. However, the tracking process is always affected by reflection noises and blocking obstacles in the environment. This paper provides a robust and optimal algorithm for tracking a moving object on the surface of water. First, we create a matrix to project the image pixels back to the real world coordinate. Second, color and shape tests are used to recognize the object and a vector is used to represent the object. If the object is partially blocked by the obstacles or the reflection from the water surface, the vector is used to predict the position of the body. In the real-time tracking, a Kalman filter is used to optimize the prediction. We tested our algorithm by tracking a submarine on the water surface of a tank. Experimental results show that the visual tracking method is robust to reflection noises and blocking obstacles. Show less