General Electric Company

Lm Wind Power A/s

Xiao Bian

David Scott Diwinsky

Bernard Patrick Bewlay

Ser Nam Lim

Arpit Jain

Shaopeng Liu

John Karigiannis

Colin James Parris

Wei Wang

Stephane Harel

Steeves Bouchard

Huan Tan

Todd William Danko

Wayne Ray Grady

Feng Xue

Masako Yamada

Yen-Liang Lin

Longyin Wen

Walter Vincent Dixon, Iii

Byron Andrew Pritchard

Yan Liu

Venkata Vijayaraghava Nalladega

Mark Richard Gilder

David Cantin

Mustafa Gokhan Uzunbas

Jean-Philippe Choiniere

Sravanthi Bondugula

Mustafa Devrim Kaba

Swaminathan Sankaranarayanan

Maxime Beaudoin Pouliot

Shourya Sarcar

Marc-Andre Marois

Peihong Zhu

Peng Chu

Gary Anthony Miller

Michael John Macdonald

The present disclosure is related to methods and systems for controlling a snake-arm robot. The method includes receiving real-time image data associated with an operating environment or a location of a workpiece from optical sensor(s) mounted on a robot head of the robot; receiving input data describing a desired pose of the robot head; computing and translating a desired displacement of the robot head; computing a position of each of the links of the snake-arm robot to follow motion of the robot head, a current position of each the links, and data required to move joints connecting the links to move the robot to the desired pose; generating movement instructions; and transmitting the movement instructions to a drive motor associated with an introduction device or controllers associated with servo-motors operably connected to joints connecting the links of the snake-arm causing the robot head to move to the desired pose.

Utilizing optical data to control operation of a snake-arm robot

A method of inspecting a component using an image retrieval module includes storing an inspection image file in a memory and identifying a region of interest in the inspection image file. The method further includes accessing a database storing image files and determining feature vectors associated with the image files. The method also includes determining a hash code for each image file based on the feature vectors and classifying a subset of image files as relevant based on the hash codes. The method further includes sorting the subset of image files based on the feature vectors and generating search results based on the sorted subset of image files. The image retrieval module includes a convolutional neural network configured to learn from the determination of the feature vectors and increase the accuracy of the image retrieval module in classifying the image files.

Inspection systems and methods including image retrieval module

A method of inspecting a component includes storing at least one inspection image file in a memory and receiving a search request associated with the at least one inspection image file. The method also includes accessing a database including a plurality of image files, comparing the hash code of the at least one inspection image file to the hash code of each image file of the plurality of image files, and identifying a first subset of image files based on the hash code comparison. The method also includes comparing the feature data of the at least one inspection image file to the feature data of each image file of the first subset of image files and classifying a second subset of image files as relevant based on the feature data comparison. The method further includes generating search results based on the second subset of image files.

A method of inspecting a component using an image inspection controller that includes a processor communicatively coupled to a memory includes classifying each sample image in a first database as a first sample or a second sample using a classification module, extracting at least one class generic feature from each first sample to generate a plurality of class generic features, and extracting at least one class specific feature from each second sample to generate a plurality of class specific features. The method further includes combining the class generic features and the class specific features to generate a plurality of supplemental images. The method further includes storing the sample images and the supplemental images in a second database, classifying each sample image and each supplemental image, capturing at least one image of the component using a camera, and identifying at least one feature of the component in the at least one image of the component using the classification module.

Inspection systems and methods including image classification module

Methods and systems for controlling a snake-arm robot. In an embodiment, a server computer receives real-time image data associated with at least one of an operating environment and a location of a workpiece from an optical sensor mounted on a robot head of a snake-arm robot, and receives, input data describing a desired pose of the robot head from a user device. The server computer then computes a desired velocity of the robot head using an image Jacobian, translates the desired velocity of the robot head into incremental displacement data and rotation data within a control cycle, computes a position of each of a plurality of links comprising a snake-arm of the snake-arm robot to follow motion of the robot head, computes a current position of each of the plurality of links utilizing a forward dynamics model, and computes force and torque data required to move at least one of a plurality of joints connecting the links to move the snake-arm robot to the desired pose. The method also includes generating movement instructions based on the force and torque data, and transmitting the movement instructions to at least one of a drive motor associated with an introduction device and a plurality of controllers associated with servo-motors operably connected to joints connecting the links of the snake arm causing the robot head to move to the desired pose.

Utilizing optical data to dynamically control operation of a snake-arm robot

A system includes a borescope and at least one processor. The borescope includes a camera configured to acquire an acquisition series of frames of at least one target component. The at least one processor is operably coupled to the camera, and is configured to acquire the acquisition series of frames from the camera; determine a blurriness metric value for each of the frames; select frames that satisfy a threshold for the blurriness metric value to form an inspection series of frames; and perform an inspection analysis for the at least one target component using the inspection series of frames.

Systems and methods for improved component inspection

An inspection system includes one or more imaging devices and one or more processors. The imaging devices generate a first set of images of a work piece at a first position relative to the work piece and a second set of images of the work piece at a second position relative to the work piece. At least some of the images in the first and second sets are acquired using different light settings. The processors analyze the first set of images to generate a first prediction image associated with the first position, and analyze the second set of images to generate a second prediction image associated with the second position. The first and second prediction images include respective candidate regions. The processors merge the first and second prediction images to detect at least one predicted defect in the work piece depicted in at least one of the candidate regions.

System and method for work piece inspection

An inspection system includes an imaging device, visible light source, ultraviolet light source, and at least one processor. The imaging device generates a first image set of a work piece while the ultraviolet light source illuminates the work piece with ultraviolet light to cause fluorescent dye thereon to emit light, and generates a second image set of the work piece while the visible light source illuminates the work piece with visible light. The first and second image sets are generated at the same positions of the imaging device relative to the work piece. The processor maps the second image set to a computer design model of the work piece based on features depicted in the second image set and the positions of the imaging device. The processor determines a defect location on the work piece based on an analysis of the first image set and the computer design model.

Fluorescent penetrant inspection system and method

An inspection system includes one or more processors that obtain a first image of a work piece that has a fluorescent dye thereon in an ultraviolet (UV) light setting and a second image of the work piece in a visible light setting. The first and second images are generated by one or more imaging devices in the same position relative to the work piece. The one or more processors identify a candidate region of the first image based on a light characteristic of one or more pixels, and determine a corresponding candidate region of the second image that is at an analogous location as the candidate region of the first image. The one or more processors analyze both candidate regions to detect a potential defect on a surface of the work piece and a location of the potential defect relative to the surface of the work piece.

A system includes one or more processors configured to detect perimeter edges of one or more rotor blades of a turbine assembly as depicted in a series of image frames using boundary analysis performed on the image frames. The one or more processors are configured to identify a set of the image frames as key frames based on positional offsets between the perimeter edges that are detected in the image frames and a reference blade pose such that the key frames are able to be inspected for objects-of-interest without inspecting the image frames that are not the key frames.

System and method for image-based target object inspection

Systems and methods are provided relating to artificial neural networks are provided. The systems and methods obtain a teacher network that includes artificial neural layers configured to automatically identify one or more objects in an image examined by the artificial neural layers, receive a set of task images at the teacher network, examine the set of task images with the teacher network, identify a subset of the artificial neural layers that are utilized during examination of the set of task images with the teacher network, and define a student network based on the set of task images. The student network is configured to automatically identify one or more objects in an image examined by the subset.

Neural network systems

A system includes one or more processors configured to analyze obtained image data representing a rotor blade to detect a candidate feature on the rotor blade and determine changes in the size or position of the candidate feature over time. The one or more processors are configured to identify the candidate feature on the rotor blade as a defect feature responsive to the changes in the candidate feature being the same or similar to a predicted progression of the defect feature over time. The predicted progression of the defect feature is determined according to an action-guidance function generated by an artificial neural network via a machine learning algorithm. Responsive to identifying the candidate feature on the rotor blade as the defect feature, the one or more processors are configured to automatically schedule maintenance for the rotor blade, alert an operator, or stop movement of the rotor blade.

Neural network feature recognition system

A generative adversarial network (GAN) system includes a generator neural sub-network configured to receive one or more images depicting one or more objects. The generator neural sub-network also is configured to generate a foreground image and a background image based on the one or more images that are received, the generator neural sub-network configured to combine the foreground image with the background image to form a consolidated image. The GAN system also includes a discriminator neural sub-network configured to examine the consolidated image and determine whether the consolidated image depicts at least one of the objects. The generator neural sub-network is configured to one or more of provide the consolidated image or generate an additional image as a training image used to train another neural network to automatically identify the one or more objects in one or more other images.

Visual anomaly detection system

A method includes obtaining a series of images of a rotating target object through multiple revolutions of the target object. The method includes grouping the images into multiple, different sets of images. The images in each of the different sets depict a common portion of the target object. At least some of the images in each set are obtained during a different revolution of the target object. The method further includes examining the images in at least a first set of the multiple sets of images using an artificial neural network for automated object-of-interest recognition by the artificial neural network.

Image analysis neural network systems

A generative adversarial network (GAN) system includes a generator sub-network configured to examine one or more images of actual damage to equipment. The generator sub-network also is configured to create one or more images of potential damage based on the one or more images of actual damage that were examined. The GAN system also includes a discriminator sub-network configured to examine the one or more images of potential damage to determine whether the one or more images of potential damage represent progression of the actual damage to the equipment.

Growing community of inventors

Niskayuna, NY, United States of America

Xiao Bian