Computer system and method for grinding chatter detection

Abstract

Systema and methods for grinding chatter detection and measurement are provided herein. A system includes a light source configured to project a light pattern on a surface to be inspected; a camera configured to collect light reflection data, the light reflection data including a modulation of the light pattern; and a processing server configured to process collected light reflection data to detect and measure grinding chatter. A method includes projecting a light pattern on a surface to be inspected; collecting light reflection data, the light reflection data including modulation of the light pattern; and processing collected light reflection data to detect and measure grinding chatter. Another method includes executing via a computer system comprising at least one processor: receiving image data; and processing the image data to detect and measure grinding chatter.

Description

COMPUTER SYSTEM AND METHOD FOR GRINDING CHATTER DETECTION Technical Field

[0001] The following relates generally to manufacturing, and more particularly to systems and methods for grinding chatter detection.Introduction

[0002] During grinding and polishing processes, tiny vibrations between a workpiece and cutting tool result in waves on the workpiece surface. These waves are known as grinding chatter. Grinding chatter occurs in almost all manufacturing processes, as heavy machines used in manufacturing generate a high amount of vibration. The presence of grinding chatter can result in a poor surface finish and a decline in dimensional accuracy.

[0003] While there may be several methods to avoid grinding chatter, if a particular workpiece has grinding chatter (or too much grinding chatter), it may be necessary to discard it. Thus, techniques to detect and measure grinding chatter are needed. Existing grinding chatter detection and measurement methods use either contact or visual inspection systems, measuring submicron surface roughness using optical systems. However, existing techniques to detect and measure grinding chatter have several disadvantages.

[0004] Firstly, they are difficult to be integrated in real-time, as parts must be separately imaged or examined to directly measure the surface roughness.

[0005] Further, current techniques often require tuning or reconfiguration for different parts in order to examine those parts. This can lead to inefficient amounts of time spent configuring different setups for different parts.

[0006] Moreover, setting up a current grinding chatter detection and measurement system can be very expensive. Thus, many manufacturers may be unable to access these technologies due to the costs involved.

[0007] Accordingly, there is a need for an improved system and method for grinding chatter detection and measurement that overcomes at least some of the disadvantages of existing systems and methods.

[0008] This background information is provided to reveal information believed by the applicant to be of possible relevance to the present disclosure. No admission is necessarily intended, nor should be construed, that any of the preceding information constitutes prior art against the present disclosure.Summary

[0009] A system for detection and measurement of grinding chatter is provided. The system includes a light source configured to project a light pattern on a surface to be inspected; a camera configured to collect light reflection data, the light reflection data including a modulation of the light pattern; and a processing server configured to process collected light reflection data to detect and measure grinding chatter.

[0010] In an embodiment, the processing server includes an analysis module.

[0011] In an embodiment, the analysis module is configured to analyze the collected light reflection data using a grinding chatter detection and measurement model.

[0012] In an embodiment, the grinding chatter detection and measurement model includes a pattern detection technique.

[0013] In an embodiment, the pattern detection technique is configured to receive the collected light reflection data as an input.

[0014] In an embodiment, the grinding chatter detection and measurement model is configured to generate a score describing grinding chatter detection and measurement as an output.

[0015] In an embodiment, the light source is at least one of: a collimated projector; a collimated laser; a laser line generator; and a collimated projector / laser with a pattern generating element.

[0016] In an embodiment, the light source is at least one of: totally coherent; partially coherent; and totally incoherent.

[0017] In an embodiment, the light source is a pattern generator made by at least one of: a refractive element; a diffractive element; and a combination of refractive and diffractive elements.

[0018] In an embodiment, the light pattern is at least one of: a line; a circle; a plurality of dots; and any combination of a line, and circle and a plurality of dots.

[0019] In an embodiment, the light pattern is oriented to traverse grinding chatter pattern perturbations.

[0020] In an embodiment, a length of the light pattern is long enough to sample a plurality of periods of grinding chatter pattern; and thin enough to avoid averaging signals from large areas of the surface combining grinding chatter patterns of variable distributions.

[0021] In an embodiment, the light reflection data is collected from at least one of: the surface; and an intermediate plane.

[0022] In an embodiment, the intermediate plane includes a diffusive material that is either transmissive or reflective.

[0023] In an embodiment, the pattern detection technique includes a modulated light pattern detection model.

[0024] In an embodiment, the score is output as at least one of: a numerical score; a binary score; and a categorical score.

[0025] A method of detection and measurement of grinding chatter is provided. The method comprising: projecting a light pattern on a surface to be inspected; collecting light reflection data, the light reflection data including modulation of the light pattern; processing collected light reflection data to detect and measure grinding chatter.

[0026] In an embodiment, the processing includes using an analysis module.

[0027] In an embodiment, the analysis module is configured to analyze the collected light reflection data using a grinding chatter detection and measurement model.

[0028] In an embodiment, the grinding chatter detection and measurement model includes a pattern detection technique.

[0029] In an embodiment, the pattern detection technique is configured to receive the collected light reflection data as an input.

[0030] In an embodiment, the grinding chatter detection and measurement model is configured to generate a score describing grinding chatter detection and measurement as an output.

[0031] In an embodiment, the light pattern is at least one of: a line; a circle, a plurality of dots; and a combination of the above.

[0032] In an embodiment, the light pattern is generated by at least one of: a collimated projector; a collimated laser; a laser line generator; and a collimated projector / laserwith a pattern generating element.

[0033] In an embodiment, the light pattern is generated by at least one of: a totally coherent light source; a partially coherent light source; and a totally incoherent light source.

[0034] In an embodiment, a pattern generator of the light pattern is made by at least one of: a refractive element; a diffractive element; and a combination of refractive and diffractive elements.

[0035] In an embodiment, the light pattern is oriented to traverse grinding chatter pattern perturbations.

[0036] In an embodiment, a length of the light pattern is long enough to sample a plurality of periods of grinding chatter pattern; and thin enough to avoid averaging signals from large areas of the surface combining grinding chatter patterns of variable distributions.

[0037] In an embodiment, the light reflection data is collected from at least one of: the surface; and an intermediate plane.

[0038] In an embodiment, the intermediate plane includes a diffusive material that is either transmissive or reflective.

[0039] In an embodiment, the pattern detection technique includes a modulated light pattern detection model.

[0040] In an embodiment, the score is output as at least one of: a numerical score; a binary score; and a categorical score.

[0041] A computer implemented method for detection and measurement of grinding chatter is provided. The method includes executing via a computer system comprising at least one processor: receiving image data; and processing the image data to detect and measure grinding chatter.

[0042] In an embodiment, the processing includes using an analysis module.

[0043] In an embodiment, the analysis module is configured to analyze the image data using a grinding chatter detection and measurement model.

[0044] In an embodiment, the grinding chatter detection and measurement model includes a pattern detection technique.

[0045] In an embodiment, the pattern detection technique is configured to receive the image data as an input.

[0046] In an embodiment, the grinding chatter detection and measurement model is configured to generate a score describing grinding chatter detection and measurement as an output.

[0047] In an embodiment, the pattern detection technique includes a modulated light pattern detection model.

[0048] In an embodiment, the score is output as at least one of: a numerical score; a binary score; and a categorical score.

[0049] Other aspects and features will become apparent, to those ordinarily skilled in the art, upon review of the following description of some exemplary embodiments.Brief Description of the Drawings

[0050] The drawings included herewith are for illustrating various examples of articles, methods, and apparatuses of the present specification. In the drawings:

[0051] Figure 1 is a schematic diagram of a system for detection and measurement of grinding chatter, according to an embodiment;

[0052] Figure 2 is a flowchart of a method of detection and measurement of grinding chatter, according to an embodiment;

[0053] Figure 3 is a flowchart of a computer implemented method for detection and measurement of grinding chatter, according to an embodiment;

[0054] Figure 4 is a schematic diagram of a device for detection and measurement of grinding chatter, according to an embodiment;

[0055] Figure 5 is a schematic diagram of an exemplary electronic device for detection and measurement of grinding chatter, according to an embodiment; and

[0056] Figure 6 is an image of example outputs and scores for detection and measurement of grinding chatter, according to an embodiment.Detailed Description

[0057] Various apparatuses or processes will be described below to provide an example of each claimed embodiment. No embodiment described below limits any claimed embodiment and any claimed embodiment may cover processes or apparatuses that differ from those described below. The claimed embodiments are not limited to apparatuses or processes having all of the features of any one apparatus or process described below or to features common to multiple or all of the apparatuses described below.

[0058] As used herein, the term “about” should be read as including variation from the nominal value, for example, a + / -10% variation from the nominal value. It is to be understood that such a variation is always included in a given value provided herein, whether or not it is specifically referred to.

[0059] One or more systems described herein may be implemented in computer programs executing on programmable computers, each comprising at least one processor, a data storage system (including volatile and non-volatile memory and / or storage elements), at least one input device, and at least one output device. For example, and without limitation, the programmable computer may be a programmable logic unit, a mainframe computer, server, and personal computer, cloud-based program or system, laptop, personal data assistance, cellular telephone, smartphone, or tablet device.

[0060] Each program is preferably implemented in a high-level procedural or object-oriented programming and / or scripting language to communicate with a computer system. However, the programs can be implemented in assembly or machine language, if desired. In any case, the language may be a compiled or interpreted language. Each such computer program is preferably stored on a storage media or a device readable by a general or special purpose programmable computer for configuring and operating the computer when the storage media or device is read by the computer to perform the procedures described herein.

[0061] A description of an embodiment with several components in communication with each other does not imply that all such components are required. On the contrary, a variety of optional components are described to illustrate the wide variety of possible embodiments of the present disclosure.

[0062] Further, although process steps, method steps, algorithms or the like may be described (in the disclosure and I or in the claims) in a sequential order, such processes, methods and algorithms may be configured to work in alternate orders. In other words, any sequence or order of steps that may be described does not necessarily indicate a requirement that the steps be performed in that order. The steps of processes described herein may be performed in any order that is practical. Further, some steps may be performed simultaneously.

[0063] When a single device or article is described herein, it will be readily apparent that more than one device I article (whether or not they cooperate) may be used in place of a single device I article. Similarly, where more than one device or article is described herein (whether or not they cooperate), it will be readily apparent that a single device I article may be used in place of the more than one device or article.

[0064] The following relates generally to manufacturing, and more particularly to systems and methods for grinding chatter detection and measurement using scattered light.

[0065] Existing techniques for detecting and measuring grinding chatter have a number of disadvantages including that they are difficult to integrate in real-time, often require tuning or reconfiguration for different parts, and are generally very expensive.

[0066] As a solution, techniques disclosed herein use an optical non-imaging method to infer grinding chatter occurrences indirectly by interpreting light reflection patterns from surfaces, without the need to directly measure the surface roughness.

[0067] Specifically, systems and methods are provided whereby a light source is used to project a light pattern on a surface of a part to be inspected, and a camera is used to collect light reflections. The collected light reflections are then processed and analyzed. The analysis may leverage artificial intelligence and machine learning systems which receive the collected light reflections as input and generate a score describing grinding chatter detection and measurementas an output. Grinding chatter occurrences may be inferred indirectly by interpreting collected light reflections.

[0068] Advantageously, the components for such techniques may only include a light source, a camera, and a computer to process the information. Further, such a system may be easily integrated into existing manufacturing processes due to the simplicity of components. Furthermore, minimal reconfiguration is required for different parts, as the principle of how the grinding chatter is inferred does not vary. Moreover, techniques disclosed herein may be substantially more cost effective to implement, as compared to known techniques.

[0069] Referring now to Figure 1, shown therein is a system 100 for detection and measurement of grinding chatter, according to an embodiment.

[0070] The system 100 includes a light source 105 configured to project a light pattern 107 on a surface 110 to be inspected.

[0071] The system 100 further includes a camera 115 configured to collect light reflection data 117, the light reflection data 117 including a reflection of the light pattern 107.

[0072] In various embodiments, a light pattern 107 hitting a perfectly smooth surface 110 will be reflected uniformly onto the camera 115 sensor creating a uniform intensity pattern.

[0073] As a result, grinding chatter, having a non-uniform surface perturbation will be mapped into a non-uniform light distribution on the camera 115 sensor, resulting in a modulated light reflection pattern that is collected by the camera 115.

[0074] The system 100 further includes a processing server 120 configured to process collected light reflection data 119 to detect and measure grinding chatter.

[0075] The processing server 120 may include an analysis module 122 configured to analyze the collected light reflection data 119.

[0076] The analysis module 122 may include a grinding chatter detection and measurement model 124 trained to detect and measure grinding chatter.

[0077] The grinding chatter detection and measurement model 124 may include a pattern detection technique 126.

[0078] In various embodiments, the pattern detection technique 126 includes a machine learning model.

[0079] In various embodiments, the pattern detection technique 126 includes a neural network model (such as, for example, a deep learning module).

[0080] In various embodiments, the analysis module 122 includes non-machine learning based vision / image processing techniques.

[0081] In various embodiments, the grinding chatter detection and measurement model 124 includes a neural network model (e.g., a deep learning model).

[0082] In various embodiments, the grinding chatter detection and measurement model 124 includes a non-neural network-based machine learning model.

[0083] In various embodiments, the grinding chatter detection and measurement model 124 includes a statistical model.

[0084] In various embodiments, the grinding chatter detection and measurement model 124 includes a rule-based decision algorithm.

[0085] In various embodiments, the grinding chatter detection and measurement model 124 includes a computer vision algorithm.

[0086] In various embodiments, the grinding chatter detection and measurement model 124 includes image processing.

[0087] In various embodiments, the grinding chatter detection and measurement model 124 includes object detection models.

[0088] In various embodiments, the grinding chatter detection and measurement model 124 includes segmentation models.

[0089] Object detection and segmentation models may be used to provide more granular analysis of the areas of the image that are indicators of identified grinding chatter.

[0090] The pattern detection technique 126 may receive the collected light reflection data 119 as an input.

[0091] The grinding chatter detection and measurement model 124 may generate a score 132 describing grinding chatter detection and measurement as an output 130.

[0092] In various embodiments, Al algorithms are used to evaluate the intensity distribution of the collected light reflection data 119. Specifically, Al algorithms trained to analyze modulated light patterns are used to evaluate patterns corresponding to grinding chatter and to provide a score 132 that is correlated to a roughness of the surface 110. Thus, a severity of the grinding chatter / roughness may be predicted indirectly without the need to perform a direct surface measurement.

[0093] In an embodiment, the light source 105 may be a collimated projector, a collimated laser, a laser line generator, a collimated projector / laser with a pattern generating element, or a combination of the above.

[0094] In an embodiment, the light source 105 may be either totally coherent, partially coherent, totally incoherent, or a combination of the above.

[0095] In an embodiment, the light source 105 may be a pattern generator made by either a refractive element, a diffractive element, or a combination of refractive and diffractive elements.

[0096] In an embodiment, the light pattern 107 may be a line, a circle, a dot, or a plurality of the above.

[0097] In an embodiment, the light pattern 107 is oriented to traverse grinding chatter pattern perturbations.

[0098] In an embodiment, a width of the light pattern 107 is long enough to sample at least a plurality of periods of grinding chatter pattern, and thin enough to avoid averaging signals from large areas of the surface 110 combining grinding chatter patterns of variable distributions.

[0099] In an embodiment, the light reflection data 117 is collected from at least one of the surface 110, and an intermediate plane 135. In an embodiment, the intermediate plane 135 includes a diffusive material that can be either transmissive or reflective.

[0100] In various embodiments, the intermediate plane 135 is imaged using the camera 115 that is focused on the intermediate plane 135.

[0101] In various embodiments, the camera 115 may receive light reflection data 117 from either the surface 110 or the intermediate plane 135.

[0102] When not using the intermediate plane 135, the camera lens is set to infinity, such that it is not imaging the surface 110 but is capturing scattered reflected light.

[0103] Advantageously, this precludes the need to refocus varying distances from a rotating surface 110 of varying heights, as the light reflection data 117 is collected by the camera 115 irrespective of minor differences in distance.

[0104] Further, as a surface 110 will generally be convex in shape, the light reflection data 117 will always be able to reach the camera 115. Advantageously, this precludes the need to know the exact shape and orientation of rotating surfaces as described by lift files that are required for grinding chatter measurement by current available techniques.

[0105] Moreover, the camera 115 and light source 105 may be mounted at close proximity to the surface 110 of the part being examined as it doesn’t need to obey the minimum working distances given by the lens equation for an imaging system. This has the added benefit of allowing for a very compact setup.

[0106] In an embodiment, the pattern detection technique 126 includes a modulated light pattern detection model.

[0107] In an embodiment, the score 132 is output as at least one of a numerical score, a binary score, and a categorical score.

[0108] In various embodiments, a single numerical score with a greater output score may correspond to elevated grinding chatter. For example, the score 132 may quantify the level of grinding chatter out of 100 (e.g., as a percentage). Similarly, the score 132 may include a binary determination of grinding chatter, with a value of “1” corresponding to grinding chatter and with a value of “0” corresponding to no grinding chatter.

[0109] In various embodiments, a binary determination may include two scores indicating the likelihood of two classes.

[0110] In various embodiments, the score 132 includes a categorical score. For example, the score 132 may be assigned from a fixed set of three or more possible categories with each corresponding to a level of grinding chatter (e.g., no grinding chatter, weak grinding chatter, or strong grinding chatter). The categorical score may be determined by converting a numerical score to a categorical score, with each category corresponding to a range of possible numerical score values.

[0111] In various embodiments, categorical scores may be obtained via multiple numerical scores whose relative values are compared, with each score indicating the likelihood of one of the categories.

[0112] In various embodiments, the score 132 includes a roughness value (e.g. out of a particular scale) associated with the surface 110. This would allow for a more granular characterization of the level of grinding chatter present.

[0113] In various embodiments, light reflection data 117 from multiple areas of a surface 110 (or multiple surfaces 110) reaches the camera 115 and appears as multiple spatially separated patterns within the camera 115 field of vision (FOV).

[0114] This characteristic may be utilized to measure different part locations (or different parts) simultaneously by projecting multiple-line patterns on multiple different areas of a surface 110 (or multiple different parts) and using a single camera 115 to collect light reflection data 117 into one image.

[0115] Further, reflected light patterns are spatially separated and do not overlap, allowing each location (or part) to be examined easily (i.e., determining which location or part may be faulty).

[0116] In various embodiments, neural network models are selected from existing (i.e., “off the shelf’) Al and machine learning architectures that are suitable for image processing.

[0117] In various embodiments, neural networks are adapted for the various score configurations by adjusting output formats and optimization objectives according to a target output.

[0118] In various embodiments, models are trained using gradient based optimization methods to optimize the network weights.

[0119] In various embodiments, gradient based optimization methods may include at least one of stochastic gradient descent (SGD), batch gradient descent, mini-batch gradient descent (with or without momentum), Root Mean Squared Propagation (RMSprop), and Adam optimization algorithm.

[0120] In various embodiments, model training may include image augmentation. Data augmentation may be performed to increase the variability of the input images, so that the model has higher robustness to the images obtained from different conditions. Two augmentations strategies that may be implemented include photometric distortions and geometric distortions in aiding with training models. In dealing with photometric distortion, the brightness, contrast, hue, saturation, and noise of an image may be adjusted. For geometric distortion, random scaling, cropping, flipping, and rotating may be added. Other special augmentations may include, for example: random erase, CutOut, Dropout / Drop Connect / Dropblock in the feature map, Mixup (with two different images), CutMix, Mosaic augmentation.

[0121] By augmenting images during training, a training set may be grown artificially, which may increase robustness of the generalizing capability of the model.

[0122] In various embodiments, models may be selected based on a variety of criteria including, but not limited to, accuracy, robustness, interpretability, training time, and scalability.

[0123] Referring now to Figure 2, shown therein is a flowchart of a method 200 of detection and measurement of grinding chatter, according to an embodiment. All or parts of the method 200 may be implemented at or by the system 100 of Figure 1.

[0124] At 210, the method 200 includes projecting a light pattern on a surface to be inspected.

[0125] At 220, the method 200 further includes collecting light reflection data, the light reflection data including a reflection of the modulated light pattern.

[0126] In various embodiments, a light pattern hitting a perfectly smooth surface will be reflected uniformly onto the camera sensor creating a uniform intensity distribution.

[0127] As a result, grinding chatter, having a non-uniform surface perturbation will be mapped into a non-uniform light distribution on a camera sensor, resulting in a modulated light reflection pattern that is collected by the camera.

[0128] At 230, the method 200 further includes processing collected light reflection data to detect and measure grinding chatter.

[0129] In an embodiment, the processing includes using an analysis module.

[0130] In an embodiment, the analysis module is configured to analyze the collected light reflection data using a grinding chatter detection and measurement model.

[0131] In an embodiment, the grinding chatter detection and measurement model includes a pattern detection technique.

[0132] In various embodiments, the pattern detection technique includes a machine learning model.

[0133] In various embodiments, the pattern detection technique includes a neural network model (such as, for example, a deep learning module).

[0134] In various embodiments, the analysis module includes non-machine learning based vision / image processing techniques.

[0135] In various embodiments, the grinding chatter detection and measurement model includes a neural network model (e.g., a deep learning model).

[0136] In various embodiments, the grinding chatter detection and measurement model includes a non-neural network-based machine learning model.

[0137] In various embodiments, the grinding chatter detection and measurement model includes a statistical model.

[0138] In various embodiments, the grinding chatter detection and measurement model includes a rule-based decision algorithm.

[0139] In various embodiments, the grinding chatter detection and measurement model includes a computer vision algorithm.

[0140] In various embodiments, the grinding chatter detection and measurement model includes image processing.

[0141] In various embodiments, the grinding chatter detection and measurement model includes object detection models.

[0142] In various embodiments, the grinding chatter detection and measurement model includes segmentation models.

[0143] Object detection and segmentation models may be used to provide more granular analysis of the areas of the image that are indicators of identified grinding chatter.

[0144] In an embodiment, the pattern detection technique is configured to receive the collected light reflection data as an input.

[0145] In an embodiment, the grinding chatter detection and measurement model generates a score describing grinding chatter detection and measurement as an output.

[0146] In various embodiments, Al algorithms are used to evaluate the intensity distribution of the collected light reflection data. Specifically, Al algorithms trained to analyze modulated light patterns are used to evaluate patterns corresponding to grinding chatter and to provide a score that is correlated to a roughness of the surface. Thus, a severity of the grinding chatter / roughness may be predicted indirectly without the need to perform a direct surface measurement.

[0147] In an embodiment, the light pattern is at least one of a line, a circle, a plurality of dots, and any combination of a line, a circle, and a plurality of dots.

[0148] In an embodiment, the light pattern is generated by at least one of a collimated projector, a collimated laser, a laser line generator, and a collimated projector / laserwith a pattern generating element.

[0149] In an embodiment, the light pattern is generated by at least one of a totally coherent light source, a partially coherent light source, and a totally incoherent light source.

[0150] In an embodiment, a pattern generator of the light pattern is made by at least one of a refractive element, a diffractive element, and a combination of refractive and diffractive elements.

[0151] In an embodiment, the light pattern is oriented to traverse grinding chatter pattern perturbations.

[0152] In an embodiment, a length of the light pattern is long enough to at least a few periods of grinding chatter pattern, and thin enough to avoid averaging a signal from large areas of the surface combining grinding chatter patterns of variable distributions.

[0153] In an embodiment, the light reflection data is collected from at least one of the surface, or an intermediate plane.

[0154] In an embodiment, the intermediate plane includes a diffusive material that is either transmissive or reflective.

[0155] In various embodiments, the intermediate plane is imaged using the camera that is focused on the intermediate plane.

[0156] When not using the intermediate plane, the camera lens is set to infinity, such that it is not imaging the surface but is capturing scattered reflected light.

[0157] In an embodiment, the pattern detection technique includes a modulated light pattern detection model.

[0158] In an embodiment, the score is output as at least one of a numerical score, a binary score, and a categorical score.

[0159] In various embodiments, a single numerical score with a greater output score may correspond to elevated grinding chatter. For example, the score may quantify the level of grinding chatter out of 100 (e.g., as a percentage). Similarly, the score may include a binary determination of grinding chatter, with a value of “1” corresponding to grinding chatter and with a value of “0” corresponding to no grinding chatter.

[0160] In various embodiments, the score includes a categorical score. For example, the score may be assigned from a fixed set of three or more possible categories with each corresponding to a level of grinding chatter (e.g., no grinding chatter, weak grinding chatter, or strong grinding chatter). The categorical score may be determined by converting a numerical score to a categorical score, with each category corresponding to a range of possible numerical score values.

[0161] In various embodiments, the score includes a roughness value (e.g. out of a particular scale) associated with the surface. This would allow for a more granular characterization of the level of grinding chatter present.

[0162] In various embodiments, light reflection data from multiple areas of a surface (or multiple surfaces) reaches the camera and appears as multiple separate patterns within the camera field of vision (FOV).

[0163] This characteristic may be utilized to measure different part locations (or different parts) simultaneously by projecting multiple-line patterns on multiple different areas of a surface (or two different parts) and using a single camera to collect light reflection data into one image.

[0164] Further, reflected light patterns are spatially separated and do not overlap allowing each location (or part) to be examined easily (i.e., determining which location or part may be faulty).

[0165] In various embodiments, neural network models are selected from existing (i.e., “off the shelf’) Al and machine learning architectures that are suitable for image processing.

[0166] In various embodiments, neural networks are adapted for the various score configurations by adjusting output formats and optimization objectives according to a target output.

[0167] In various embodiments, models are trained using stochastic gradient descent to optimize the network weights.

[0168] Referring now to Figure 3, shown therein is a flowchart of a computer implemented method 300 for detection and measurement of grinding chatter, according to an embodiment. All or parts of the method 300 may be implemented at or by the system 100 of Figure 1.

[0169] At 310, the method 300 includes receiving image data.

[0170] At 320, the method 300 further includes processing the image data to detect and measure grinding chatter.

[0171] In various embodiments, processing the image data includes reading the image data.

[0172] In various embodiments, processing the image data includes preprocessing the image data.

[0173] In various embodiments, preprocessing the image data includes normalizing the image data by shifting and rescaling RGB values.

[0174] In various embodiments, preprocessing the image data includes converting the image data to greyscale.

[0175] In various embodiments, preprocessing the image data includes cropping the image data.

[0176] In various embodiments, preprocessing the image data includes applying a zero padding technique to the image data.

[0177] In various embodiments, preprocessing the image data includes extracting a region of interest (ROI) of the image data.

[0178] In various embodiments, preprocessing the image data includes filtering the image data.

[0179] In various embodiments, processing the image data includes using a model (e.g., a machine learning model, a neural network model, etc.).

[0180] In various embodiments, processing the image data includes generating an output.

[0181] In various embodiments, processing the image data includes sending the output to an output device.

[0182] In various embodiments, processing the image data includes performing an action based on the output.

[0183] In an embodiment, the processing includes using an analysis module.

[0184] In an embodiment, the analysis module is configured to analyze the image data using a grinding chatter detection and measurement model.

[0185] In an embodiment, the grinding chatter detection and measurement model includes a pattern detection technique.

[0186] In an embodiment, the pattern detection technique is configured to receive the image data as an input.

[0187] In an embodiment, the grinding chatter detection and measurement model is configured to generate a score describing grinding chatter detection and measurement as an output.

[0188] In an embodiment, the pattern detection technique includes a modulated light pattern detection model.

[0189] In an embodiment, the score is output as at least one of a numerical score, a binary score, and a categorical score.

[0190] In various embodiments, a single numerical score with a greater output score may correspond to elevated grinding chatter. For example, the score may quantify the level of grinding chatter out of 100 (e.g., as a percentage). Similarly, the score may include a binary determination of grinding chatter, with a value of “1” corresponding to grinding chatter and with a value of “0” corresponding to no grinding chatter.

[0191] In various embodiments, the score includes a categorical score. For example, the score may be assigned from a fixed set of three or more possible categories with each corresponding to a level of grinding chatter (e.g., no grinding chatter, weak grinding chatter, or strong grinding chatter). The categorical score may be determined by converting a numerical score to a categorical score, with each category corresponding to a range of possible numerical score values.

[0192] In various embodiments, the score includes a roughness value (e.g. out of a particular scale) associated with the surface. This would allow for a more granular characterization of the level of grinding chatter present.

[0193] Referring now to Figure 4, shown therein is a schematic diagram of a device 400 for detection and measurement of grinding chatter, according to an embodiment.

[0194] The device 400 includes a network interface 405 and processing electronics 410.

[0195] The network interface 405 may be a single network interface or a combination of network interfaces. The network interface 405 may include an optical communication interface or radio communication interface, such as a transmitter and receiver.

[0196] The processing electronics 410 may include a computer processer executing program instructions stored in memory, or other electronics components such as digital circuitry, including for example FPGAs and ASICs.

[0197] The device 400 may include several functional components, each of which is partially or fully implemented using the underlying network interface 405 and processing electronics 410.

[0198] The device 400 further includes a light source 415 configured to project a light pattern on a surface to be inspected. The light source 415 may include a collimated projector, and a laser line generator.

[0199] The device 400 further includes a camera 420 configured to collect light reflection data, the light reflection data including a reflection of the light pattern projected by the light source 415.

[0200] The device further includes an inspection system 425 configured to process collected light reflection data to detect and measure grinding chatter.

[0201] In various embodiments, the inspection system 425 may be an automated inspection system (AIS).

[0202] In various embodiments, the device 400 may serve as a sensing package in a larger system. In some embodiments, the device 400 may be a standalone automated visual inspection system within a manufacturing line or quality lab.

[0203] Referring now to Figure 5, shown therein is a schematic diagram of an electronic device 500 that may perform any or all of operations of the above methods and features explicitly or implicitly described herein, according to different embodiments of the present disclosure. For example, a computer equipped with network function may be configured as electronic device 500. The electronic device 500 may be used to implement the device 400 of Figure 4, for example.

[0204] As shown, the device 500 includes a processor 510, such as a Central Processing Unit (CPU) or specialized processors such as a Graphics Processing Unit (GPU) or other such processor unit, memory 520, non-transitory mass storage 530, I / O interface 540, network interface 550, and a transceiver 560, all of which are communicatively coupled via bi-directional bus 570.

[0205] According to certain embodiments, any or all of the depicted elements may be utilized, or only a subset of the elements. Further, the device 500 may contain multiple instances of certain elements, such as multiple processors, memories, or transceivers. Also, elements of the hardware device may be directly coupled to other elements without the bi-directional bus. Additionally or alternatively to a processor and memory, other electronics, such as integrated circuits, may be employed for performing the required logical operations.

[0206] The memory 520 may include any type of non-transitory memory such as static random-access memory (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), read-only memory (ROM), any combination of such, or the like.

[0207] The mass storage element 530 may include any type of non-transitory storage device, such as a solid-state drive, hard disk drive, a magnetic disk drive, an optical disk drive,USB drive, or any computer program product configured to store data and machine executable program code.

[0208] According to certain embodiments, the memory 520 or mass storage 530 may have recorded thereon statements and instructions executable by the processor 510 for performing any of the aforementioned method operations described above.

[0209] Referring now to Figure 6, shown therein is an image 600 of example outputs and scores for detection and measurement of grinding chatter, according to an embodiment.

[0210] In various embodiments, a single numerical score with a greater output score may correspond to elevated grinding chatter. For example, the score may quantify the level of grinding chatter out of 100 (e.g., as a percentage). Similarly, the score may include a binary determination of grinding chatter, with a value of “1” corresponding to grinding chatter and with a value of “0” corresponding to no grinding chatter. Numerical and binary scores may further influence an output class indicating the presence or absence of grinding chatter.

[0211] In various embodiments, the score includes a categorical score. For example, the score may be assigned from a fixed set of three or more possible categories with each corresponding to a level of the grinding chatter (e.g., no grinding chatter, weak grinding chatter, or strong grinding chatter). The categorical score may be determined by converting a numerical score to a categorical score, with each category corresponding to a range of possible numerical score values. Such categorical scores may further influence an output class indicating the overall level of grinding chatter present (e.g., strong grinding chatter).

[0212] In various embodiments, the score includes a roughness value (e.g. out of a particular scale) associated with the surface. This would allow for a more granular characterization of the level of grinding chatter present.

[0213] In various embodiments, a model outputs a single value which estimates the level of grinding chatter as measured by the surface roughness. This could also be any other measurable value that can be used to quantify grinding chatter levels.

[0214] While the above description provides examples of one or more apparatus, methods, or systems, it will be appreciated that other apparatus, methods, or systems may be within the scope of the claims as interpreted by one of skill in the art.

[0215] Elements of each embodiment may be incorporated into other embodiments, for example, configurations discussed in relation to one embodiment, may be applied to other embodiments disclosed herein.

[0216] Further, it is evident that various modifications and combinations can be made without departing from the invention. The specification and drawings are, accordingly, to be regarded simply as an illustration of the invention as defined by the claims, and are contemplatedto cover any and all modifications, variations, combinations or equivalents that fall within thescope of the present disclosure.

Claims

Claims:

1. A system for detection and measurement of grinding chatter, the system comprising:a light source configured to project a light pattern on a surface to be inspected;a camera configured to collect light reflection data, the light reflection data including a modulation of the light pattern; anda processing server configured to process collected light reflection data to detect and measure grinding chatter.

2. The system of claim 1 , wherein the processing server includes an analysis module.

3. The system of claim 2, wherein the analysis module is configured to analyze the collected light reflection data using a grinding chatter detection and measurement model.

4. The system of claim 3, wherein the grinding chatter detection and measurement model includes a pattern detection technique.

5. The system of claim 4, wherein the pattern detection technique is configured to receive the collected light reflection data as an input.

6. The system of claim 3, wherein the grinding chatter detection and measurement model is configured to generate a score describing grinding chatter detection and measurement as an output.

7. The system of claim 1 , wherein the light source is at least one of: a collimated projector;a collimated laser; a laser line generator; and a collimated projector / laser with a pattern generating element.

8. The system of claim 1 , wherein the light source is at least one of: totally coherent; partially coherent; and totally incoherent.

9. The system of claim 1, wherein the light source is a pattern generator made by at least one of: a refractive element; a diffractive element; and a combination of refractive and diffractive elements.

10. The system of claim 1 , wherein the light pattern is at least one of: a line; a circle; a plurality of dots; and any combination of a line, and circle and a plurality of dots.

11. The system of claim 1 , wherein the light pattern is oriented to traverse grinding chatter pattern perturbations.

12. The system of claim 1, wherein a length of the light pattern is long enough to sample a plurality of periods of grinding chatter pattern; and thin enough to avoid averaging signals from large areas of the surface combining grinding chatter patterns of variable distributions.

13. The system of claim 1, wherein the light reflection data is collected from at least one of:the surface; and an intermediate plane.

14. The system of claim 12, wherein the intermediate plane includes a diffusive material that is either transmissive or reflective.

15. The system of claim 4, wherein the pattern detection technique includes a modulated light pattern detection model.

16. The system of claim 6, wherein the score is output as at least one of: a numerical score;a binary score; and a categorical score.

17. A method of detection and measurement of grinding chatter, the method comprising:projecting a light pattern on a surface to be inspected;collecting light reflection data, the light reflection data including modulation of the light pattern;processing collected light reflection data to detect and measure grinding chatter.

18. The method of claim 17, wherein the processing includes using an analysis module.

19. The method of claim 18, wherein the analysis module is configured to analyze the collected light reflection data using a grinding chatter detection and measurement model.

20. The method of claim 19, wherein the grinding chatter detection and measurement model includes a pattern detection technique.

21. The method of claim 20, wherein the pattern detection technique is configured to receive the collected light reflection data as an input.

22. The method of claim 19, wherein the grinding chatter detection and measurement model is configured to generate a score describing grinding chatter detection and measurement as an output.

23. The method of claim 17, wherein the light pattern is at least one of: a line; a circle, a plurality of dots; and a combination of the above.

24. The method of claim 17, wherein the light pattern is generated by at least one of: a collimated projector; a collimated laser; a laser line generator; and a collimated projector / laser with a pattern generating element.

25. The method of claim 17, wherein the light pattern is generated by at least one of: a totally coherent light source; a partially coherent light source; and a totally incoherent light source.

26. The method of claim 17, wherein a pattern generator of the light pattern is made by at least one of: a refractive element; a diffractive element; and a combination of refractive and diffractive elements.

27. The method of claim 17, wherein the light pattern is oriented to traverse grinding chatter pattern perturbations.

28. The method of claim 17, wherein a length of the light pattern is long enough to sample a plurality of periods of grinding chatter pattern; and thin enough to avoid averaging signals from large areas of the surface combining grinding chatter patterns of variable distributions.

29. The method of claim 17, wherein the light reflection data is collected from at least one of:the surface; and an intermediate plane.

30. The method of claim 29, wherein the intermediate plane includes a diffusive material that is either transmissive or reflective.

31. The method of claim 20, wherein the pattern detection technique includes a modulated light pattern detection model.

32. The method of claim 22, wherein the score is output as at least one of: a numerical score; a binary score; and a categorical score.

33. A computer implemented method for detection and measurement of grinding chatter, the method comprising executing via a computer system comprising at least one processor:receiving image data; andprocessing the image data to detect and measure grinding chatter.

34. The method of claim 33, wherein the processing includes using an analysis module.

35. The method of claim 34, wherein the analysis module is configured to analyze the image data using a grinding chatter detection and measurement model.

36. The method of claim 35, wherein the grinding chatter detection and measurement model includes a pattern detection technique.

37. The method of claim 36, wherein the pattern detection technique is configured to receive the image data as an input.

38. The method of claim 35, wherein the grinding chatter detection and measurement model is configured to generate a score describing grinding chatter detection and measurement as an output.

39. The method of claim 36, wherein the pattern detection technique includes a modulated light pattern detection model.

40. The method of claim 38, wherein the score is output as at least one of: a numerical score;a binary score; and a categorical score.

Images