Patent Grant
12020217
This application relates generally to systems for detecting vehicle damage and estimating repair costs of such damage.
Vehicles used for transporting people and goods (such as cars, trucks, vans, etc.) can become damaged (e.g., due to collisions, vandalism, acts of nature, etc.). In such cases, it may be important to understand the estimated cost of repairing such damage to the vehicle. For example, an insurance company may need to understand this estimated cost in order to determine whether to repair the damage to the vehicle or whether to total the vehicle out. In other cases, this information may be useful to a layperson with little to no experience with vehicle damage repair cost estimation to help them understand what an expected cost of repair should be (so that they can make informed decisions about having repairs done, such as, for example, verifying that a quoted cost for making a repair under consideration is reasonable).
Current methods may rely on manual and/or in-person evaluation by a person in order to come up with such repair estimates. This may involve the services of a person who is skilled in making such evaluations. However, such persons may not be readily available at all times, and/or may not be readily available to all entities who might be interested in such information in all situations (e.g., in the case where an owner of the damaged vehicle has not involved, e.g., an insurance adjuster to analyze the damage). Accordingly, it is of value to develop ways of estimating such repair costs in a manner that does not require such manual and/or in-person evaluation.
To easily identify the discussion of any particular element or act, the most significant digit or digits in a reference number refer to the figure number in which that element is first introduced.
Machine learning methods may be developed which can augment or replace the use of manual and/or in-person evaluations of vehicle damage repair costs. These machine learning methods may train one or more neural network (NN) models in order to generate a model corresponding to the neural network that is fit to perform a given task.
One example of such a NN model may be a damage detection NN model trained to receive one or more preprocessed digital images of vehicle damage. Such preprocessed digital images may be generated from, for example, a user provided digital image (including stills from a digital video) that might be taken for purposes of use with a system including the damage detection NN model. In some cases, these user provided digital images may be taken by a layperson using a camera that is found on a smartphone (or other device). In other cases, these user provided digital images may be taken by, for example, a camera operated by a service center computer system. Other cases are contemplated. In whatever case, these user provided digital images may be unannotated (meaning that they have not been manually marked or otherwise indicated to have one or more parameters of interest preparatory to using the image to train a damage detection NN model—as may likely be the case when receiving a user provided digital image for processing at an already-trained damage detection NN model). The damage detection NN model may be trained to process a such a user provided digital image (after preprocessing) such that one or more parameters of interest regarding vehicle damage within the user provided digital image (such as a location of the vehicle damage, one or more repair parts corresponding to the vehicle damage, a type of the vehicle damage, and/or an intensity of the vehicle damage) are identified.
Another such example of a NN model may be a repair cost estimation NN model that is trained to receive one or more parameters of interest regarding vehicle damage, such as one or more repair parts corresponding to the vehicle damage, a type of vehicle damage, and an intensity of vehicle damage. This repair cost estimation NN model may be trained to process this data such that an estimation of a cost to repair such damage is made.
The use of two such trained NN models together may allow a user of a system implementing or using such models to provide an (unannotated) user provided digital image picturing, and to ultimately receive an estimated cost to repair such damage (also described herein as a “repair cost estimation”) and/or a view of segmentation data overlaid on the user provided digital image that indicates the location of the corresponding damage.
Embodiments herein discuss systems and methods for training and/or using such NN models.
It is anticipated that the application of systems leveraging these NN models may be useful in many different contexts. For example, service inspection and/or repair providers may be able to more quickly determine the nature of any damage to a vehicle and/or the cost of fixing such damage, making these service providers more efficient. As a further example, rental car companies may be able to determine the nature of damage to a vehicle both before it is rented and after it is returned in a more efficient and remote manner, allowing them to more easily integrate, for example, contactless vehicle rental methods into their business models. As a further example, vehicle insurers may be able to remotely and more quickly review damage to a vehicle and determine the cost of a claim and/or review a vehicle's current damage prior to insuring it. As a further example, laypeople that are untrained in damage analysis may use such systems to determine a reasonable expectation of the cost of making a repair under consideration.
It should be understood that such systems might provide vehicle damage and/or repair cost information via an application programming interface (API) to any consuming software system. Examples of such consuming software systems may include, for example, a vehicle repair order system (in order to capture damage of the vehicle and related estimated cost for the repair order), or a vehicle rental system (in order to, e.g., validate no damage has been done during the rental period), or any other consuming software system that the data might be useful to. These consuming software systems may be examples of “user devices” as described herein.
It is contemplated that the digital image preparation system 102 could be part of a larger computer-implemented system. Such a larger computer-implemented system could include, in addition to the digital image preparation system 102, one or more of, for example, a vehicle damage detection training system 302, a repair cost estimation training system 502, and/or a repair cost estimation system 702 as those systems are described herein.
The digital image preparation system 102 may include a memory 104, one or more processor(s) 106, one or more I/O device(s) 108, and a network interface 110. These elements may be connected by a data bus 112.
The I/O device(s) 108 may include devices connected to the digital image preparation system 102 that allow a user of the digital image preparation system 102 to provide input from the digital image preparation system 102 and receive output from the digital image preparation system 102. For example, these devices may include a mouse, a keyboard, speakers, a monitor, external storage, etc.
The network interface 110 may connect to a network 114 to allow the digital image preparation system 102 to communicate with outside entities, such as, for example, a digital image database 116, an annotated digital image database 118, and/or a vehicle damage detection system 120. Such communications may be performed via one or more APIs used by the digital image preparation system 102.
The memory 104 may contain digital image preparation instructions 122, which may be used by the processor(s) 106 to operate a digital image preparation engine 128. The memory 104 may further include (not yet annotated) digital images 124 (which may have been sourced from the digital image database 116) which are processed by the digital image preparation system 102, after which they become annotated digital images 126 (which may ultimately be stored at the annotated digital image database 118).
The method 200 includes receiving 202 digital images picturing vehicle damage from a data source. The data source could be a memory on a device implementing the method 200, or the data source could be some data source accessible to (e.g., via a network) the device implementing the method 200 (such as, for example, the digital image database 116 of
The method 200 further includes converting 204 the digital images to grey scale. This may be done prior to using the digital images in the training process so as to reduce the complexity of the analysis, in that black and white images can be represented using a single matrix, where color images may instead use multiple matrices to be represented. This may increase the training speed as compared to using color images. Converting 204 the digital images to grey scale may occur prior to annotating the digital image (see below). This may increase the accuracy of a damage detection NN model that takes as input grey scaled images (such as at least some embodiments of the damage detection NN model 320 to be described). The converting 204 the digital images to grey scale prior to training may be optional, but may be beneficial in that it leads to a decrease in the time and/or computational complexity of training the damage detection NN model.
The method 200 further includes determining 206 whether a damage detection NN model for identifying one or more parameters of interest (e.g., that are the subject of the annotation process) in an (unannotated) digital image is (already) available to the system implementing the method 200. For example, the digital image preparation engine 128 of
Labeling 212 the digital images with a damage location may be done manually (e.g., by a user of the digital image preparation system 102 using the I/O device(s) 108). The damage location may be annotated directly on the digital image in a graphical manner (e.g., using user-placed markings relative to the digital image). It is contemplated that at least some digital images may depict damage in more than one location on the vehicle. In these cases, some or all of these damage locations may be so labeled. The output of this process may be saved in annotation data corresponding to the digital image.
The method 200 then proceeds to labeling 214 the digital images with a damage type. This may be done manually (e.g., by a user of the digital image preparation system 102 using the I/O device(s) 108). The damage type used in the labeling 214 may be one of a group of pre-determined damage types for which the user is interested in training a damage detection NN model to recognize. These damage types may include (but are not limited to) a dent damage type, a scratch damage type, a smash damage type, a crack damage type, a lamp broken damage type, and a break damage type. It is contemplated that other damage types could also be used, should the user be interested in training the damage detection NN model to identify such damage types. The damage type may correspond to a labeled damage location within a digital image. In cases where digital images have more than one labeled damage location, a (potentially different) damage type may correspond to each labeled damage location. The output of this process may be saved in annotation data corresponding to the digital image.
The method 200 then proceeds to labeling 216 the digital images with a damage intensity. This may be done manually (e.g., by a user of the digital image preparation system 102 using the I/O device(s) 108). The damage intensity used in the labeling 216 may be one of a group of pre-determined damage intensities for which the user is interested in training a damage detection NN model to recognized. These damage intensities may include (but are not limited to) a minor intensity, a medium intensity, and a severe intensity. It is contemplated that other damage intensities could also be used, should the user be interested in training the damage detection NN model to identify such damage intensities. The damage intensity may correspond to a labeled damage location within a digital image. In cases where digital images have more than one labeled damage location, a (potentially different) damage intensity may correspond to each labeled damage location. The output of this process may be saved in annotation data corresponding to the digital image.
The method 200 then proceeds to labeling 218 the digital images with a repair part. This may be done manually (e.g., by a user of the digital image preparation system 102 using the I/O device(s) 108). This repair part may be a part used to effect a repair of the damaged location. In cases where digital images have more than one labeled damage location, a (potentially different) repair part may correspond to each labeled damage location. It is also anticipated that a user could specify that multiple repair parts correspond to the damage, in cases where multiple repair parts would be needed to repair such damage. The output of this process may be saved in annotation data corresponding to the digital image. The method 200 then proceeds to cleaning 220 irrelevant digital images.
Each of the labeling 212 the digital images with a damage location, the labeling 214 the digital images with a damage type, the labeling 216 the digital images with a damage intensity, and/or the labeling 218 the digital images with a repair part may be considered a form of annotation of the respective digital image. In other words, each digital image now has a set of corresponding annotation data that is to be used during the training of a damage detection NN model using the digital image. Accordingly, at this point in the method 200, it may be said that the digital images are now annotated digital images (digital images for which corresponding annotation data has been generated).
If, at determining 206, it is instead determined that a damage detection NN model for identifying one or more parameters of interest in a digital image is available, the method 200 proceeds to providing 208 the digital images to the damage detection NN model. The damage detection NN model may implement its current training to label damage location(s) and corresponding damage type(s), intensit(ies), and repair part(s) corresponding to each digital image. These labels may then be placed in annotation data corresponding to each digital image for further use during training of a damage detection NN model using each respective digital image, which may be considered a form of annotation of such digital images. Accordingly, at this point in the method 200, it may be said that the digital images are now annotated digital images.
The method 200 then proceeds to the manually verifying and/or correcting 210 of these model-annotated digital images. This may be done by a user of the digital image preparation system 102 using the I/O device(s) 108. The manually verifying and/or correcting 210 may correct any mistakes made in the annotations made by the damage detection NN model used so that the eventual training process using such the annotated digital images is more accurate. Further, if the damage detection NN model to be trained is the same as the damage detection NN model used in providing 208 the digital images to the damage detection NN model, modifications to such annotations may be made during the process of manually verifying and/or correcting 210 (or else the eventual training will simply result in the same damage detection NN model that these digital images were provided to). The method 200 then proceeds to cleaning 220 irrelevant digital images.
When cleaning 220 the irrelevant digital images, any digital image with corresponding annotation data that lacks, for example, one or more of a damage location, a damage type, a damage intensity, and/or a repair part may be removed from the set of digital images.
The method 200 further includes normalizing 222 the digital images. The normalization process may account for the fact that different digital images may have different overall ranges of values (e.g., one grey scaled digital image may have darker blacks or lighter whites than another digital image within the set). The normalization process may remove this variance from the set of digital images by causing each of values of each matrix representing the (perhaps grey scaled, as discussed above) digital image to be normalized according to a defined range (e.g., from −1 to 1). This may increase the accuracy of the eventual accuracy of a damage detection NN model that takes as input normalized images (such as at least some embodiments of the damage detection NN model 320 to be described).
The processes of converting 204 a digital image to grey scale and normalizing 222 such a digital image (whether one or both of these processes are applied) may either individually or jointly be considered a “preprocessing” (as that term is understood herein) of the digital images. Accordingly, in embodiments of the method 200 where the digital images have been normalized (and possibly grey scaled prior to normalization) at this stage, they may be considered “preprocessed digital images.” In this specification, when both a grey scaling and normalization of a digital image are discussed as being performed on a (singular) “digital image” to generate a “preprocessed digital image,” it may be understood that such processes are applied in step-wise fashion starting with the digital image, with an intervening image existing prior to completing the preprocessed digital image (but not explicitly discussed)).
The method 200 further includes augmenting 224 the set of digital images. The set of digital images may be augmented by taking one or more of the digital images and modifying it such that the modified digital image is also represented within the set. A digital image may be so modified by rotation, scaling, inversion, mirroring, cropping, etc. (with corresponding changes made to its corresponding annotation data). In this manner, a more robust data set for training may be generated.
During each of the cleaning 220 the irrelevant digital images, the normalizing 222 the digital images, and the augmenting 224 the set of digital images, the corresponding annotation data for each image may be accordingly protected/correspondingly modified. This may in some cases be handled by, for example, the digital image preparation system 102 of
The method 200 further includes saving 226 the annotated digital images. In other words, the preprocessed digital images along with their corresponding annotation data are saved as annotated digital images. These annotated digital images may be saved as, for example, the annotated digital images 126 in the memory 104 of the digital image preparation system 102 of
It is contemplated that the vehicle damage detection training system 302 could be part of a larger computer-implemented system. Such a larger computer-implemented system could include, in addition to the vehicle damage detection training system 302, one or more of, for example, the digital image preparation system 102, the repair cost estimation training system 502, and/or the repair cost estimation system 702 as those systems are described herein.
The vehicle damage detection training system 302 may include a memory 304, one or more processor(s) 306, one or more I/O device(s) 308, and a network interface 310. These elements may be connected by a data bus 312.
The I/O device(s) 308 may include devices connected to the vehicle damage detection training system 302 that allow a user of the vehicle damage detection training system 302 to provide input from the vehicle damage detection training system 302 and receive output from the vehicle damage detection training system 302. For example, these devices may include a mouse, a keyboard, speakers, a monitor, external storage, etc.
The network interface 310 may connect to a network 314 to allow the vehicle damage detection training system 302 to communicate with outside entities, such as, for example, the annotated digital image database 118. Such communications may be performed using one or more APIs used by the vehicle damage detection training system 302.
The memory 304 may contain damage detection NN model training instructions 316, which may be used by the processor(s) 306 to operate a damage detection NN model training engine 318. The memory 304 may further include a damage detection NN model 320, which may be a NN model that can be used to (and/or that is being trained to) determine a location of vehicle damage pictured in a user provided digital image, one or more repair parts corresponding to such vehicle damage; a type of the vehicle damage; and an intensity of the vehicle damage. The memory 304 may further include annotated digital images 322 that are being used by the vehicle damage detection training system 302 to train the damage detection NN model 320. In some embodiments, the annotated digital images 322 may have been sourced from, for example, the annotated digital image database 118.
The method 400 includes receiving 402 annotated digital images. The set of annotated digital images received may be provided from an outside source, such as, for example, the annotated digital image database 118. Annotated digital images so received could also be sourced from, for example, an image preparation system (such as the digital image preparation system 102 of
The method 400 further includes bifurcating 404 the set of annotated digital images into a training subset and a validation subset. The training subset may be used to train a damage detection NN model, and the validation subset may be used to check against the damage detection NN model so trained in order to make a determination regarding the accuracy of the damage detection NN model.
The method 400 further includes defining 406 a learning rate for the training process. The learning rate controls the amount of adjustment that is made for one or more neurons of the damage detection NN model in relation to a calculated loss between the output of the NN model as currently arranged and the ideal result (e.g., how accurately the damage detection NN model as currently arranged can properly identify damage location(s) and corresponding damage type(s), intensit(ies), and repair part(s)). These losses may be determinable by the system by comparing the result of using an (annotated) digital image with the damage detection NN model as currently arranged and comparing the result to the annotation data corresponding to the digital image.
The method 400 further includes defining 408 an epoch for use with the training data, which indicates how many runs through the training subset are performed during the training process.
The method 400 further includes training 410 the damage detection NN model as a mask region-based convolutional neural network (mask RCNN) using the training subset according to the learning rate and the epoch. Training the damage detection NN model as a mask RCNN may have certain advantages in embodiments herein. It may be that mask RCNNs are particularly suited for detecting one or more types of objects within a digital image. During training as a mask RCNN, the damage detection NN model may be trained to recognize both the location of and one or more classifications of an object found in a digital image using the training subset of the set of annotated digital images, which contains annotated digital images showing vehicle damage and having corresponding annotation data for each of the classification variables. For instance, a damage detection NN model trained as a mask RCNN according to embodiments herein may be capable of identifying the location of vehicle damage. This information may be returned as a mask of the vehicle damage and/or a bounding box surrounding the vehicle damage (as will be illustrated in embodiments below). Further, training the damage detection NN model as a mask RCNN may allow the damage detection NN model to classify such recognized damage according to a type (e.g., scratch, smash, dent, crack, lamp broken, etc.); according to an intensity (e.g., minor, medium, or severe etc.); and according to the corresponding repair part(s) needed to repair the pictured damage.
Transfer learning may be leveraged during the training process. For example, a known NN model may already have various trained layers that can identify points, then lines, then shapes, etc., but may not have one or more final layers that can identify damage locations, types, intensities, and/or corresponding repair parts. In these cases, the layers of the (separate) known NN model for, for example, points, lines, shapes, etc., may be re-used for the current damage detection NN model, and the training process may then train these one or more final layers for making the ultimate identifications desired (e.g., locations, types, intensities, and/or corresponding repair parts). Transfer learning leveraged in this manner may save training time.
The trained damage detection NN model (e.g., the damage detection NN model 320) may accordingly receive as input a preprocessed user provided digital image (e.g., a version of a user provided digital image that has been, for example, normalized (possibly after being grey scaled, where the preprocessed image may be grey scaled in the case where the damage detection NN model was trained on grey scaled images)). The damage detection NN model then may process such an image in order to identify the location(s) of vehicle damage in the user provided digital image. The damage detection NN model may represent and/or store these locations as segmentation data that is configured to be overlaid on the user provided digital image (e.g., configured to be overlaid on the version of the user provided digital image that is not grey scaled and/or normalized). Further, the damage detection NN model may further process the preprocessed user provided digital image to identify damage type(s) corresponding to each damage location, intensit(ies) of damage corresponding to each damage location, and/or repair part(s) needed at each damage location to effectuate a repair of the identified damage.
It is anticipated that, to the extent that annotations involving multiple repair parts for a single instance of vehicle damage are available in the digital images in the training data used to train the damage detection NN model, the damage detection NN model may be able to identify multiple repair parts that correspond to a single instance of vehicle damage.
Further, the damage detection NN model may be trained on sufficient digital images such that an appropriate accuracy is reached for determining these locations, types, intensities, and repair parts without the use of a reference image of an undamaged vehicle. In other words, it is anticipated that no reference image of an undamaged vehicle is necessary to perform methods disclosed herein using a trained damage detection NN model.
The method 400 further includes determining 412 the accuracy of the model using the validation subset. Once all epochs of training data have been run through, the damage detection NN model is run with the validation subset, and an overall loss (or average loss) based on losses corresponding to each annotated digital images of the validation subset as applied to the damage detection NN model is calculated. For example, these losses may be calculated based on how accurately the damage detection NN model as currently arranged can properly identify damage location(s) and corresponding damage type(s), intensit(ies), and repair part(s) for the digital images in the validation subset, as compared to annotation data corresponding each of the annotated digital images in the validation subset. In some embodiments, a multi-task loss function is used. For example:
Loss=Lcls+Lbox+Lmask, where:
The method 400 further includes determining 414 whether the accuracy is acceptable. If the overall loss (or average loss) calculated in relation to the validation subset is less than (or less than or equal to) a given value, the method 400 may deem that the damage detection NN model that has been trained is sufficiently accurate, and may proceed to saving 416 the trained damage detection NN model for later use.
However, if the overall loss (or average loss) calculated in relation to the validation subset is greater than (or greater than or equal to) the given value, the method 400 instead proceeds to determining 418 whether hypertuning is needed. Hypertuning may refer to changing one or both of a learning rate and/or an epoch used by the method 400 in order to generate a different resulting trained NN model. For example, in
If it is determined that hypertuning is not needed, the method 400 then concludes that additional (or new) training using additional annotated digital images may be needed in order to generate an accurate damage detection NN model. In this case, the method 400 proceeds to receiving 402 an additional and/or different set of annotated digital images such that the (now different) set may be used in conjunction with the method 400, with the hope of being able to generate a more accurate damage detection NN model with the help of this additional data.
It is contemplated that determining 418 whether hypertuning is needed and/or determining 420 that additional annotated digital images are needed to help generate a more accurate damage detection NN model could happen in any order (as both actions may be independently capable of changing/improving the output of the damage detection NN model trained by the method 400). Further, in some embodiments, it is also contemplated that perhaps only one or the other of the determining 418 whether hypertuning is needed and/or the determining 420 that additional annotated digital images are needed to help generate a more accurate damage detection NN model is performed during an iteration of the method 400.
It is contemplated that the repair cost estimation training system 502 could be part of a larger computer-implemented system. Such a larger computer-implemented system could include, in addition to the repair cost estimation training system 502, one or more of, for example, the digital image preparation system 102, the vehicle damage detection training system 302, and/or the repair cost estimation system 702 as those systems are described herein.
The repair cost estimation training system 502 may include a memory 504, one or more processor(s) 506, one or more I/O device(s) 508, and a network interface 510. These elements may be connected by a data bus 512.
The I/O device(s) 508 may include devices connected to the repair cost estimation training system 502 that allow a user of the repair cost estimation training system 502 to provide input from the repair cost estimation training system 502 and receive output from the repair cost estimation training system 502. For example, these devices may include a mouse, a keyboard, speakers, a monitor, external storage, etc.
The network interface 510 may connect to a network 514 to allow the repair cost estimation training system 502 to communicate with outside entities, such as, for example, a historical repair cost time series database 516. Such communications may be performed using one or more APIs used by the repair cost estimation training system 502.
The memory 504 may contain a repair cost estimation NN model training instructions 518, which may be used by the processor(s) 506 to operate a repair cost estimation NN model training engine 520. The memory 504 may further include a repair cost estimation NN model 522, which may be a NN model that can be used to (e.g., that is being trained to) determine an estimated cost to repair damage based on a received damage type, a received damage intensity, and one or more received repair parts. In some embodiments, a historical repair cost time series 524 may have been sourced from, for example, the historical repair cost time series database 516.
The method 600 includes initializing 602 the neuron weights within the repair cost estimation NN model that is to be trained. In transfer learning cases, these weights may be partially transferred in from another NN model.
The method 600 includes collecting 604 historical time series data for repair cost according to repair part(s), damage type, and damage intensity. The historical time series data collected may be gathered from an outside source, such as, for example, the historical repair cost time series database 516 of
The method 600 further includes standardizing 606 the historical time series data. Because the historical time series data points may be associated with items that do not have similar ranges (for example, an item corresponding to the cost of making a repair may have a range from 0 to 50,000 (dollars), where an item corresponding to the damage intensity may be a discrete classification system (minor, medium, severe)), it may be beneficial to standardize or normalize all of the items relative to each other, by removing the mean and scaling to unit variance. For example, each item may be translated (as necessary) to represent a numerical value (e.g., in the case of a damage intensity that uses the discrete classification system discussed above, minor may be translated to 1, middle may be translated to 2, and severe may be translated to 3). Then, the values corresponding to each item may be brought into a known range relative to the possible ranges for that item. For example, each data point for each item may be brought into the range of −1 to 1, while keeping the (relative distance between each separate data point intact). This known range (whether it be −1 to 1 or some other known range) may be the same range used for each item, which may act to simplify the training of the repair cost estimation NN model because differing ranges do not have to be accounted for within the training process.
The method 600 further includes bifurcating 608 the historical time series data into a training subset and a validation subset. The training subset may be used to train a repair cost estimation NN model, and the validation subset may be used to check against the repair cost estimation NN model so trained in order to make a determination regarding the accuracy of the repair cost estimation NN model.
The method 600 further includes defining 610 a learning rate for the training process. The learning rate controls the amount of adjustment that is made for one or more neurons of the repair cost estimation NN model in relation to a calculated loss between the output of the repair cost estimation NN model as currently arranged and the ideal result (e.g., how accurately the repair cost estimation NN model as currently arranged can properly estimate a cost of a repair). These losses may be determinable by the system by comparing the result of using a data point containing repair part(s), a damage type, and a damage intensity with the repair cost estimation NN model as currently arranged and comparing the resulting output of a repair cost to the actual repair cost associated with the data point.
The method 600 further includes defining 612 an epoch for use with the training data, which indicates how many runs through the training subset are performed during the training process.
The method 600 further includes using 614 a long short-term memory (LSTM) recurrent neural network (RNN) to train the repair cost estimation NN model using the training subset, according to the learning rate and the epoch.
Training the repair cost estimation NN model as an LSTM RNN may have certain advantages in embodiments herein. It may be that LSTM RNNs are particularly suited for predicting a result based on a input set of data with a view of analogous data collected over a previous time period. LSTM RNNs may be able to make predictions of results that account for, at least in part, variations over time within such data (and, e.g., in response to trends over time) Accordingly, during training as an LSTM RNN, the repair cost estimation NN model may be trained to make a repair cost estimation based on repair part(s) corresponding to vehicle damage, a type of the vehicle damage, and an intensity of the vehicle damage and in view of pricing trends over time using the training subset of the historical time series data, which contains this type of data for each timestamped data point. This information may be returned as textual information (as illustrated below).
In some embodiments, it is contemplated that a repair cost estimation NN model (e.g., the repair cost estimation NN model 522) can be trained to receive other types of inputs. For example, the data points of the training subset of the historical time series data may also include a geographic indication (such as city, zip code, country, county, etc.) where the repair took place. The training process may then configure the repair cost estimation NN model 522 to account for geographical location (as well) when making a repair cost estimation.
The trained repair cost estimation NN model (e.g., the repair cost estimation NN model 522) may accordingly receive as input a damage type of vehicle damage, a damage intensity of the vehicle damage, and repair part(s) used to repair the vehicle damage. The repair cost estimation NN model then may process this information and identify an estimated cost to repair the vehicle damage at the current time.
In some embodiments, it is contemplated that a repair cost estimation NN model (e.g., the repair cost estimation NN model 522) can give output corresponding to additional types of inputs. For example, the repair cost estimation NN model 522 may be trained to receive a geographic indication (as described above) such that it can further determine the estimated cost to repair damage based on a received geographic indication.
The method 600 further includes determining 616 the accuracy of the model using the validation subset. Once all epochs of training data have been run through, the repair cost estimation NN model is run with the validation subset, and an overall loss (or average loss) based on losses corresponding to each data point of the validation subset as applied to the repair cost estimation NN model is calculated. For example, these losses may be calculated based on how accurately the repair cost estimation NN model as currently arranged can properly estimate a cost of repair at the timestamped time for each data point in the validation subset using the repair part(s) corresponding to the vehicle damage, the type of the vehicle damage, and the intensity of the vehicle damage from each respective data point, as compared to the actual repair cost at that same time that is known for that data point.
The method 600 further includes determining 618 whether the accuracy is acceptable. If the overall loss (or average loss) calculated in relation to the validation subset is less than (or less than or equal to) a given value, the method 600 may deem that the repair cost estimation NN model that has been trained is sufficiently accurate, and may proceed to determining 620 the trained repair cost estimation NN model for later use.
However, if the overall loss (or average loss) calculated in relation to the validation subset is greater than (or greater than or equal to) the given value, the method 600 instead proceeds to determining 620 whether hypertuning is needed. Hypertuning may refer to changing one or both of a learning rate and/or an epoch used by the method 600 in order to generate a different resulting trained repair cost estimation NN model. For example, in
If it is determined that hypertuning is not needed, the method 600 then concludes that additional or different historical time series data may be needed in order to generate an accurate model. In this case, the method 600 proceeds to collecting 604 (additional) historical time series data such that a larger and/or different set of such data may be used in conjunction with the method 600, with the hope of being able to generate a more accurate repair cost estimation NN model with the help of this additional data.
It is contemplated that determining 620 whether hypertuning is needed and/or determining 622 that additional historical time series data are needed to help generate a more accurate repair cost estimation NN model could happen in any order (as both actions may be independently capable of changing/improving the output of the repair cost estimation NN model trained by the method 600). Further, in some embodiments, it is also contemplated that perhaps only one or the other of determining 620 whether hypertuning is needed and/or determining 622 that additional historical time series data is needed to help generate a more accurate repair cost estimation NN model is performed during an iteration of the method 600.
It is contemplated that the repair cost estimation system 702 could be part of a larger computer-implemented system. Such a larger computer-implemented system could include, in addition to the repair cost estimation system 702, one or more of, for example, the digital image preparation system 102, the vehicle damage detection training system 302, and/or the repair cost estimation training system 502 as those systems are described herein.
The repair cost estimation system 702 may include a memory 704, one or more processor(s) 706, one or more I/O device(s) 708, and a network interface 710. These elements may be connected by a data bus 712.
The I/O device(s) 708 may include devices connected to the repair cost estimation system 702 that allow a user of the repair cost estimation system 702 to provide input from the repair cost estimation system 702 and receive output from the repair cost estimation system 702. For example, these devices may include a mouse, a keyboard, speakers, a monitor, external storage, etc.
The network interface 710 may connect to a network 714 to allow the repair cost estimation system 702 to communicate with outside entities, such as, for example, a user device 716, a repair parts pricing database 718, the vehicle damage detection training system 302, and/or the repair cost estimation training system 502. The repair parts pricing database 718 may be a database containing current information regarding the price(s) of one or more repair parts. The vehicle damage detection training system 302 may be, e.g., the vehicle damage detection training system 302 of
The memory 704 may contain repair cost estimation instructions 720, which may be used by the processor(s) 706 to operate a repair cost estimation engine 722. The memory 704 may further include the damage detection NN model 320, which may be a NN model that has been trained to identify a location of, type of, intensity of, and repair part(s) corresponding to one or more instances of pictured vehicle damage in a preprocessed user provided digital image, as described above. The damage detection NN model 320 may have been received at the repair cost estimation system 702 from the vehicle damage detection training system 302 after training at the vehicle damage detection training system 302, in the manner described above. The memory 704 may further include the repair cost estimation NN model 522 which may be a NN model that has been trained to determine an estimated cost to repair damage based on a received damage type, a received damage intensity, and received repair part(s) to make a repair, in the manner described above. This repair cost estimation NN model 522 may have been received at the repair cost estimation system 702 from the repair cost estimation training system 502 after training at the repair cost estimation training system 502, in the manner described above. The memory 704 may further include a user provided digital image 724. The user provided digital image 724 may have been received from the user device 716. The user provided digital image 724 may depict a damaged vehicle that has one or more damage locations, types, and/or intensities, in the manner described herein. The memory 704 may further include a preprocessed user provided digital image 726 that was generated by the repair cost estimation system 702 from the user provided digital image 724 using the processor(s) 706. The memory 704 may further include results data 728 that contains the results of processing at the processor(s) 706 of the preprocessed user provided digital image 726 with the damage detection NN model 320. This results data 728 may include one or more of segmentation data to be overlaid on the user provided digital image 724 indicating the location of the vehicle damage, an estimated cost to repair the vehicle damage, the type of the vehicle damage, the intensity of the vehicle damage, and/or the one or more repair part(s) corresponding to the vehicle damage.
The method 800 includes receiving 802 a user provided digital image containing unannotated vehicle damage. For example, the repair cost estimation system 702 may receive such user provided digital image as the user provided digital image 724 over the network 714 from the user device 716.
The method 800 further includes generating 804 a preprocessed user provided digital image. This may be done by applying normalization to the user provided digital image. This may further be accomplished in some cases by applying grey scaling to the user provided digital image, which may allow compatibility with the damage detection NN model 320, in the case the damage detection NN model 320 was trained using grey scaled digital images, in the manner described above. This generating 804 may be performed by the processor(s) 706 using the repair cost estimation engine 722. This generating 804 may generate the preprocessed user provided digital image 726.
The method 800 further includes providing 806, to a damage detection NN model, the preprocessed user provided digital image. For example, the repair cost estimation system 702 may provide the preprocessed user provided digital image 726 to the damage detection NN model 320 for processing at the repair cost estimation system 702 (using the one or more processor(s) 706). Alternatively, the preprocessed user provided digital image 726 may be provided over the network 714 to a damage detection NN model hosted at the vehicle damage detection training system 302, which then processes the preprocessed user provided digital image 726 using the hosted damage detection NN model 320 using one or more processor(s) of the vehicle damage detection training system 302.
The method 800 further includes receiving 808, from the damage detection NN model, a location of the damage, corresponding repair part(s), a type of the damage, and an intensity of the damage. In the case of processing that occurs on the vehicle damage detection training system 302, the vehicle damage detection training system 302 returns these results to the repair cost estimation system 702. These results may represent the location of damage in the user provided digital image 724 as segmentation data that is configured to be overlaid on the user provided digital image 724. Further, these results may identify a damage type corresponding to the damage location, an intensity of the damage corresponding to the damage location, and/or one or more repair parts needed at the damage location to effectuate a repair of the identified damage. These results may be stored in the results data 728.
The method 800 further includes providing 810, to a repair cost estimation NN model, the corresponding repair part(s), the type of the damage, and the intensity of the damage. For example, the repair cost estimation system 702 may provide this information to the repair cost estimation NN model 522 for processing at the repair cost estimation system 702 (using the processor(s) 706). Alternatively, this information may be provided over the network 714 to a repair cost estimation NN model hosted at the repair cost estimation training system 502, which then processes this data using the hosted repair cost estimation NN model 522 using one or more processor(s) of the repair cost estimation training system 502. This information may be sourced from the results data 728 (as previously there stored).
The method 800 further includes receiving 812, from the repair cost estimation NN model, an estimated repair cost. This result may represent the expected or estimated total cost of repairing the unannotated vehicle damage from the user provided digital image at the present time. This result may be stored in the results data 728.
The method 800 further optionally includes determining 814 a current cost of the corresponding repair part(s). The current cost of the corresponding repair part(s) may be knowable in real time by communicating with, for example, the repair parts pricing database 718.
The method 800 further optionally includes updating 816 the estimated repair cost based on the current cost of the corresponding repair part(s). This may involve adjusting the estimated repair cost based on how well the current cost of the corresponding repair part(s) is in line with the repair cost estimation. For example, if the current cost of (any one of) the corresponding repair part(s) is determined to be much higher or lower recently (e.g., as discovered by referring again to the repair parts pricing database 718), the repair cost estimation system 702 may determine that the repair part costs that were used to train the repair cost estimation NN model used may have been somewhat out of date, and may adjust the estimated repair cost upward or downward accordingly.
The method 800 further includes providing 818, to a user device, results data comprising the estimated repair cost and segmentation data overlaid on the user provided digital image indicating the location of the unannotated vehicle damage. For example, the estimated repair cost may be an estimated repair cost predicted by the repair cost estimation NN model 522 and stored at the results data 728. The segmentation data may also be sourced from the results data 728. Then, such segmentation data is overlaid by the repair cost estimation system 702 on the user provided digital image 724. The estimated repair cost and the user provided digital image 724 overlaid with the segmentation data are then delivered to the user device 716 over the network 714.
As shown, first vehicle damage segmentation 1006 may illustrate the location of the first vehicle damage 1012, the second vehicle damage segmentation 1008 may illustrate the location of the second vehicle damage 1014, and the third vehicle damage segmentation 1010 may illustrate the location of third vehicle damage 1016. These illustrations may be shown by using the illustrated bounding boxes and/or masks showing the location of the respective identified damage. Further, as illustrated, each of these bounding boxes may include for display a textual representation of a damage type corresponding to the damage pictured. Additional and/or alternative textual representations may be similarly made, as discussed in further detail below.
Further, the first vehicle damage 1012, the second vehicle damage 1014, and the third vehicle damage 1016 may have each been independently identified by the repair cost estimation system 702 as being of one of various possible damage types. For example, the first vehicle damage 1012 may have been identified by the repair cost estimation system 702 as a scratch damage type, the second vehicle damage 1014 may have been identified by the repair cost estimation system 702 as a dent damage type, and the third vehicle damage 1016 may have been identified by the repair cost estimation system 702 as a smash damage type. As illustrated, these types may respectively be displayed textually in, for example, the first vehicle damage segmentation 1006, the second vehicle damage segmentation 1008, and the third vehicle damage segmentation 1010.
Further, the first vehicle damage 1012, the second vehicle damage 1014, and the third vehicle damage 1016 may have each been independently identified by the repair cost estimation system 702 as being of various possible damage intensities. For example, the first vehicle damage 1012 may have been identified as of a severe damage intensity, the second vehicle damage 1014 may have been identified by the repair cost estimation system 702 as of a minor damage intensity, and the third vehicle damage 1016 may have been identified by the repair cost estimation system 702 as of a severe damage intensity. While not illustrated, it is contemplated that these intensities may respectively be displayed textually in, for example, the first vehicle damage segmentation 1006, the second vehicle damage segmentation 1008, and the third vehicle damage segmentation 1010.
Various (and potentially different) repair parts may have been identified by the repair cost estimation system 702 as corresponding to the first vehicle damage 1012, the second vehicle damage 1014, and the third vehicle damage 1016. For example, a “rear passenger door” repair part may have been identified to correspond to the first vehicle damage 1012, a “none” repair part may have been identified to correspond to the second vehicle damage 1014 (meaning that the repair can be done without a repair part), and a “front passenger door” repair part may have been identified to correspond to the third vehicle damage 1016. It is further anticipated that, to the extent that data involving multiple repair parts for a single instance of vehicle damage is available in the training data used to train the damage detection NN model 320, the repair cost estimation system 702 may be able to identify multiple repair parts that correspond to one of the first vehicle damage 1012, the second vehicle damage 1014, and/or the third vehicle damage 1016. For example, in addition to the “front passenger door” part, a “front passenger window mechanism” may also be identified as corresponding to the third vehicle damage 1016. While not illustrated, it is contemplated that these repair parts may respectively be displayed textually in, for example, the first vehicle damage segmentation 1006, the second vehicle damage segmentation 1008, and the third vehicle damage segmentation 1010.
Once identified, the damage type, the damage intensity, and the corresponding repair part(s) for one or more of the first vehicle damage 1012, the second vehicle damage 1014, and/or the third vehicle damage 1016 may be processed with the repair cost estimation NN model 522 in order to generate a repair cost estimation for one (or more) of these. This repair cost estimation may also be an example of results data that is sent to and displayed on the user device 716. While not illustrated, it is contemplated that the repair cost estimations may respectively be displayed textually in, for example, the first vehicle damage segmentation 1006, the second vehicle damage segmentation 1008, and the third vehicle damage segmentation 1010.
It is further contemplated that the damage type, the damage intensity, and the corresponding repair part(s) for each of these could also be sent to (and displayed on) the user device 716 as results data.
As shown, first vehicle damage segmentation 1106 may illustrate the location of the first vehicle damage 1114, the second vehicle damage segmentation 1108 may illustrate the location of the second vehicle damage 1116, the third vehicle damage segmentation 1110 may illustrate the location of the third vehicle damage 1118, and the fourth vehicle damage segmentation 1112 may illustrate the location of the fourth vehicle damage 1120. These illustrations may be shown by using the illustrated bounding boxes and/or masks showing the location of the respective identified damage. Further, as illustrated, each of these bounding boxes may include for display a textual representation of a damage type corresponding to the damage pictured. Additional and/or alternative textual representations may be similarly made, as discussed in further detail below.
Further, the first vehicle damage 1114, the second vehicle damage 1116, the third vehicle damage 1118, and the fourth vehicle damage 1120 may have each been independently identified by the repair cost estimation system 702 as being of one of various possible damage types. For example, the first vehicle damage 1114 may have been identified by the repair cost estimation system 702 as a lamp broken damage type, the second vehicle damage 1116 may have been identified by the repair cost estimation system 702 as a crack damage type, the third vehicle damage 1118 may have been identified by the repair cost estimation system 702 as a smash damage type, and the fourth vehicle damage 1120 may have been identified by the repair cost estimation system 702 as a scratch damage type. As illustrated, these types may respectively be displayed textually in, for example, the first vehicle damage segmentation 1106, the second vehicle damage segmentation 1108, the third vehicle damage segmentation 1110, and the fourth vehicle damage segmentation 1112.
Further, the first vehicle damage 1114, the second vehicle damage 1116, the third vehicle damage 1118, and the fourth vehicle damage 1120 may have each been independently identified by the repair cost estimation system 702 as being of various possible damage intensities. For example, the first vehicle damage 1114 may have been identified as of a medium damage intensity, the second vehicle damage 1116 may have been identified by the repair cost estimation system 702 as of a severe damage intensity, the third vehicle damage 1118 may have been identified by the repair cost estimation system 702 as of a severe damage intensity, and the fourth vehicle damage 1120 may have been identified by the repair cost estimation system 702 as of a medium damage intensity. While not illustrated, it is contemplated that these intensities may respectively be displayed textually in, for example, the first vehicle damage segmentation 1106, the second third vehicle damage segmentation 1110, the third vehicle damage segmentation 1110, and the fourth vehicle damage segmentation 1112.
Various (and potentially different) repair parts may have been identified by the repair cost estimation system 702 as corresponding to the first vehicle damage 1114, the second vehicle damage 1116, the third vehicle damage 1118, and the fourth vehicle damage 1120. For example, a “front driver headlamp” repair part may have been identified to correspond to the first vehicle damage 1114, a “front bumper” repair part may have been identified to correspond to the second vehicle damage 1116, a “front driver side panel” repair part may have been identified to correspond to the third vehicle damage 1118, and a “front driver side panel” repair part may have been identified to correspond to the fourth vehicle damage 1120. It is further anticipated that, to the extent that data involving multiple repair parts for a single instance of vehicle damage is available in the training data used to train the damage detection NN model 320, the repair cost estimation system 702 may be able to identify multiple repair parts that correspond to one of the first vehicle damage 1114, the second vehicle damage 1116, the third vehicle damage 1118, and/or the fourth vehicle damage 1120. For example, in addition to the identification of the “front bumper” repair part, a “radiator” repair part may also be identified as corresponding to the second vehicle damage 1116. While not illustrated, it is contemplated that these repair parts may respectively be displayed textually in, for example, the first vehicle damage segmentation 1106, the second third vehicle damage segmentation 1110, the third vehicle damage segmentation 1110, and the fourth vehicle damage segmentation 1112.
Once identified, the damage type, the damage intensity, and the corresponding repair part(s) for one or more of the first vehicle damage 1114, the second vehicle damage 1116, the third vehicle damage 1118, and/or the fourth vehicle damage 1120 may be processed with the repair cost estimation NN model 522 in order to generate a repair cost estimation for one (or more) of these. This repair cost estimation may also be an example of results data that is sent to and displayed on the user device 716. It is further contemplated that the damage type, the damage intensity, and the corresponding repair part(s) for each of these could also be sent to (and displayed on) the user device 716 as results data.
The user device 716 may display some of the results data in a display region 1202. The results data shown in the display region 1202 may include the user provided digital image 1002 that was provided to the repair cost estimation system 702 as described in relation to
The user device 716 may display some of the results data in a textual information region 1204. In the example of
Textual information region 1204 may also include textual information for the second vehicle damage 1014 and the third vehicle damage 1016, and a user may cause an indicator similar to the indicator 1208 to appear in relation to either of these within the display region 1202 by making a selection of the associated corresponding textual information.
The foregoing specification has been described with reference to various embodiments, including the best mode. However, those skilled in the art appreciate that various modifications and changes can be made without departing from the scope of the present disclosure and the underlying principles of the present disclosure. Accordingly, this disclosure is to be regarded in an illustrative rather than a restrictive sense, and all such modifications are intended to be included within the scope thereof. Likewise, benefits, other advantages, and solutions to problems have been described above with regard to various embodiments. However, benefits, advantages, solutions to problems, and any element(s) that may cause any benefit, advantage, or solution to occur or become more pronounced are not to be construed as a critical, required, or essential feature or element.
As used herein, the terms “comprises,” “comprising,” or any other variation thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus.
Embodiments herein may include various engines, which may be embodied in machine-executable instructions to be executed by a general-purpose or special-purpose computer (or other electronic device). Alternatively, the engine functionality may be performed by hardware components that include specific logic for performing the function(s) of the engines, or by a combination of hardware, software, and/or firmware.
Principles of the present disclosure may be reflected in a computer program product on a tangible computer-readable storage medium having stored instructions thereon that may be used to program a computer (or other electronic device) to perform processes described herein. Any suitable computer-readable storage medium may be utilized, including magnetic storage devices (hard disks, floppy disks, and the like), optical storage devices (CD-ROMs, DVDs, Blu-Ray discs, and the like), flash memory, and/or other types of medium/machine readable medium suitable for storing electronic instructions. These instructions may be loaded onto a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions that execute on the computer or other programmable data processing apparatus create means for implementing the functions specified. These instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified. The instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified.
Principles of the present disclosure may be reflected in a computer program implemented as one or more software modules or components. As used herein, a software module or component may include any type of computer instruction or computer-executable code located within a memory device and/or computer-readable storage medium. A software module may, for instance, comprise one or more physical or logical blocks of computer instructions, which may be organized as a routine, a program, an object, a component, a data structure, etc., that perform one or more tasks or implement particular data types.
In certain embodiments, a particular software module may comprise disparate instructions stored in different locations of a memory device, which together implement the described functionality of the module. Indeed, a module may comprise a single instruction or many instructions, and may be distributed over several different code segments, among different programs, and across several memory devices. Some embodiments may be practiced in a distributed computing environment where tasks are performed by a remote processing device linked through a communications network. In a distributed computing environment, software modules may be located in local and/or remote memory storage devices. In addition, data being tied or rendered together in a database record may be resident in the same memory device, or across several memory devices, and may be linked together in fields of a record in a database across a network.
Suitable software to assist in implementing the disclosed embodiments is readily provided by those of skill in the pertinent art(s) using the teachings presented here and programming languages and tools, such as Java, JavaScript, Pascal, C++, C, database languages, APIs, SDKs, assembly, firmware, microcode, and/or other languages and tools.
Embodiments as disclosed herein may be computer-implemented in whole or in part on a digital computer. The digital computer includes a processor performing the required computations. The computer further includes a memory in electronic communication with the processor to store a computer operating system. The computer operating systems may include, but are not limited to, MS-DOS, Windows, Linux, Unix, AIX, CLIX, QNX, OS/2, and MacOS. Alternatively, it is expected that future embodiments will be adapted to execute on other future operating systems.
It will be obvious to those having skill in the art that many changes may be made to the details of the above-described embodiments without departing from the underlying principles of the disclosed embodiments.
| Number | Name | Date | Kind |
|---|---|---|---|
| 3792445 | Bucks et al. | Feb 1974 | A |
| 4258421 | Juhasz et al. | Mar 1981 | A |
| 4992940 | Dworkin | Feb 1991 | A |
| 5003476 | Abe | Mar 1991 | A |
| 5034889 | Abe | Jul 1991 | A |
| 5058044 | Stewart et al. | Oct 1991 | A |
| 5421015 | Khoyi et al. | May 1995 | A |
| 5442553 | Parrillo | Aug 1995 | A |
| 5452446 | Johnson | Sep 1995 | A |
| 5521815 | Rose, Jr. | May 1996 | A |
| 5649186 | Ferguson | Jul 1997 | A |
| 5694595 | Jacobs et al. | Dec 1997 | A |
| 5729452 | Smith et al. | Mar 1998 | A |
| 5764943 | Wechsler | Jun 1998 | A |
| 5787177 | Leppek | Jul 1998 | A |
| 5790785 | Klug et al. | Aug 1998 | A |
| 5835712 | Dufresne | Nov 1998 | A |
| 5845299 | Arora et al. | Dec 1998 | A |
| 5862346 | Kley et al. | Jan 1999 | A |
| 5911145 | Arora et al. | Jun 1999 | A |
| 5956720 | Fernandez et al. | Sep 1999 | A |
| 5974149 | Leppek | Oct 1999 | A |
| 5974418 | Blinn et al. | Oct 1999 | A |
| 5974428 | Gerard et al. | Oct 1999 | A |
| 5978776 | Seretti et al. | Nov 1999 | A |
| 5987506 | Carter et al. | Nov 1999 | A |
| 6003635 | Bantz et al. | Dec 1999 | A |
| 6006201 | Berent et al. | Dec 1999 | A |
| 6009410 | Lemole et al. | Dec 1999 | A |
| 6018748 | Smith | Jan 2000 | A |
| 6021416 | Dauerer et al. | Feb 2000 | A |
| 6021426 | Douglis et al. | Feb 2000 | A |
| 6026433 | D'Arlach et al. | Feb 2000 | A |
| 6041310 | Green et al. | Mar 2000 | A |
| 6041344 | Bodamer et al. | Mar 2000 | A |
| 6055541 | Solecki et al. | Apr 2000 | A |
| 6061698 | Chadha et al. | May 2000 | A |
| 6067559 | Allard et al. | May 2000 | A |
| 6070164 | Vagnozzi | May 2000 | A |
| 6134532 | Lazarus et al. | Oct 2000 | A |
| 6151609 | Truong | Nov 2000 | A |
| 6178432 | Cook et al. | Jan 2001 | B1 |
| 6181994 | Colson et al. | Jan 2001 | B1 |
| 6185614 | Cuomo et al. | Feb 2001 | B1 |
| 6189104 | Leppek | Feb 2001 | B1 |
| 6216129 | Eldering | Apr 2001 | B1 |
| 6219667 | Lu et al. | Apr 2001 | B1 |
| 6236994 | Schwartz et al. | May 2001 | B1 |
| 6240365 | Bunn | May 2001 | B1 |
| 6263268 | Nathanson | Jul 2001 | B1 |
| 6285932 | De Belledeuille et al. | Sep 2001 | B1 |
| 6289382 | Bowman-Amuah | Sep 2001 | B1 |
| 6295061 | Park et al. | Sep 2001 | B1 |
| 6330499 | Chou et al. | Dec 2001 | B1 |
| 6343302 | Graham | Jan 2002 | B1 |
| 6353824 | Boguraev et al. | Mar 2002 | B1 |
| 6356822 | Diaz et al. | Mar 2002 | B1 |
| 6374241 | Lamburt et al. | Apr 2002 | B1 |
| 6397226 | Sage | May 2002 | B1 |
| 6397336 | Leppek | May 2002 | B2 |
| 6401103 | Ho et al. | Jun 2002 | B1 |
| 6421733 | Tso et al. | Jul 2002 | B1 |
| 6473849 | Keller et al. | Oct 2002 | B1 |
| 6496855 | Hunt et al. | Dec 2002 | B1 |
| 6505106 | Lawrence et al. | Jan 2003 | B1 |
| 6505205 | Kothuri et al. | Jan 2003 | B1 |
| 6519617 | Wanderski et al. | Feb 2003 | B1 |
| 6529948 | Bowman-Amuah | Mar 2003 | B1 |
| 6535879 | Behera | Mar 2003 | B1 |
| 6539370 | Chang et al. | Mar 2003 | B1 |
| 6546216 | Mizoguchi et al. | Apr 2003 | B2 |
| 6553373 | Boguraev et al. | Apr 2003 | B2 |
| 6556904 | Larson et al. | Apr 2003 | B1 |
| 6564216 | Waters | May 2003 | B2 |
| 6571253 | Thompson et al. | May 2003 | B1 |
| 6581061 | Graham | Jun 2003 | B2 |
| 6583794 | Wattenberg | Jun 2003 | B1 |
| 6594664 | Estrada et al. | Jul 2003 | B1 |
| 6606525 | Muthuswamy et al. | Aug 2003 | B1 |
| 6629148 | Ahmed et al. | Sep 2003 | B1 |
| 6640244 | Bowman-Amuah et al. | Oct 2003 | B1 |
| 6643663 | Dabney et al. | Nov 2003 | B1 |
| 6654726 | Hanzek | Nov 2003 | B1 |
| 6674805 | Kovacevic et al. | Jan 2004 | B1 |
| 6678706 | Fishel | Jan 2004 | B1 |
| 6697825 | Underwood et al. | Feb 2004 | B1 |
| 6701232 | Yamaki | Mar 2004 | B2 |
| 6721747 | Lipkin | Apr 2004 | B2 |
| 6728685 | Ahluwalia | Apr 2004 | B1 |
| 6738750 | Stone et al. | May 2004 | B2 |
| 6744735 | Nakaguro | Jun 2004 | B1 |
| 6748305 | Klausner et al. | Jun 2004 | B1 |
| 6785864 | Te et al. | Aug 2004 | B1 |
| 6795819 | Wheeler et al. | Sep 2004 | B2 |
| 6823258 | Ukai et al. | Nov 2004 | B2 |
| 6823359 | Heidingsfeld | Nov 2004 | B1 |
| 6826594 | Pettersen | Nov 2004 | B1 |
| 6847988 | Toyouchi et al. | Jan 2005 | B2 |
| 6850823 | Eun et al. | Feb 2005 | B2 |
| 6871216 | Miller et al. | Mar 2005 | B2 |
| 6901430 | Smith | Mar 2005 | B1 |
| 6894601 | Grunden et al. | May 2005 | B1 |
| 6917941 | Wight et al. | Jul 2005 | B2 |
| 6922674 | Nelson | Jul 2005 | B1 |
| 6941203 | Chen | Sep 2005 | B2 |
| 6944677 | Zhao | Sep 2005 | B1 |
| 6954731 | Montague et al. | Oct 2005 | B1 |
| 6963854 | Boyd et al. | Nov 2005 | B1 |
| 6965806 | Eryurek et al. | Nov 2005 | B2 |
| 6965968 | Touboul | Nov 2005 | B1 |
| 6978273 | Bonneau et al. | Dec 2005 | B1 |
| 6981028 | Rawat et al. | Dec 2005 | B1 |
| 6990629 | Heaney et al. | Jan 2006 | B1 |
| 6993421 | Pillar | Jan 2006 | B2 |
| 7000184 | Matveyenko et al. | Feb 2006 | B2 |
| 7003476 | Samra et al. | Feb 2006 | B1 |
| 7010495 | Samra et al. | Mar 2006 | B1 |
| 7028072 | Kliger et al. | Apr 2006 | B1 |
| 7031554 | Iwane | Apr 2006 | B2 |
| 7039704 | Davis et al. | May 2006 | B2 |
| 7047318 | Svedloff | May 2006 | B1 |
| 7062343 | Ogushi et al. | Jun 2006 | B2 |
| 7062506 | Taylor et al. | Jun 2006 | B2 |
| 7072943 | Landesmann | Jul 2006 | B2 |
| 7092803 | Kapolka et al. | Aug 2006 | B2 |
| 7107268 | Zawadzki et al. | Sep 2006 | B1 |
| 7124116 | Huyler | Oct 2006 | B2 |
| 7152207 | Underwood et al. | Dec 2006 | B1 |
| 7155491 | Schultz et al. | Dec 2006 | B1 |
| 7171418 | Blessin | Jan 2007 | B2 |
| 7184866 | Squires et al. | Feb 2007 | B2 |
| 7197764 | Cichowlas | Mar 2007 | B2 |
| 7219234 | Ashland et al. | May 2007 | B1 |
| 7240125 | Fleming | Jul 2007 | B2 |
| 7246263 | Skingle | Jul 2007 | B2 |
| 7281029 | Rawat | Oct 2007 | B2 |
| 7287000 | Boyd et al. | Oct 2007 | B2 |
| 7322007 | Schowtka et al. | Jan 2008 | B2 |
| 7386786 | Davis et al. | Jun 2008 | B2 |
| 7401289 | Lachhwani et al. | Jul 2008 | B2 |
| 7406429 | Salonen | Jul 2008 | B2 |
| 7433891 | Haber et al. | Oct 2008 | B2 |
| 7457693 | Olsen et al. | Nov 2008 | B2 |
| 7477968 | Lowrey | Jan 2009 | B1 |
| 7480551 | Lowrey et al. | Jan 2009 | B1 |
| 7496543 | Bamford et al. | Feb 2009 | B1 |
| 7502672 | Kolls | Mar 2009 | B1 |
| 7536641 | Rosenstein et al. | May 2009 | B2 |
| 7548985 | Guigui | Jun 2009 | B2 |
| 7587504 | Adams et al. | Sep 2009 | B2 |
| 7590476 | Shumate | Sep 2009 | B2 |
| 7593925 | Cadiz et al. | Sep 2009 | B2 |
| 7593999 | Nathanson | Sep 2009 | B2 |
| 7613627 | Doyle et al. | Nov 2009 | B2 |
| 7620484 | Chen | Nov 2009 | B1 |
| 7624342 | Matveyenko et al. | Nov 2009 | B2 |
| 7657594 | Banga et al. | Feb 2010 | B2 |
| 7664667 | Ruppelt et al. | Feb 2010 | B1 |
| 7739007 | Logsdon | Jun 2010 | B2 |
| 7747680 | Ravikumar et al. | Jun 2010 | B2 |
| 7778841 | Bayer et al. | Aug 2010 | B1 |
| 7801945 | Geddes et al. | Sep 2010 | B1 |
| 7818380 | Tamura et al. | Oct 2010 | B2 |
| 7861309 | Spearman et al. | Dec 2010 | B2 |
| 7865409 | Monaghan | Jan 2011 | B1 |
| 7870253 | Muilenburg et al. | Jan 2011 | B2 |
| 7899701 | Odom | Mar 2011 | B1 |
| 7908051 | Oesterling | Mar 2011 | B2 |
| 7979506 | Cole | Jul 2011 | B2 |
| 8010423 | Bodin et al. | Aug 2011 | B2 |
| 8019501 | Breed | Sep 2011 | B2 |
| 8036788 | Breed | Oct 2011 | B2 |
| 8051159 | Muilenburg et al. | Nov 2011 | B2 |
| 8055544 | Ullman et al. | Nov 2011 | B2 |
| 8060274 | Boss et al. | Nov 2011 | B2 |
| 8095403 | Price | Jan 2012 | B2 |
| 8099308 | Uyeki | Jan 2012 | B2 |
| 8135804 | Uyeki | Mar 2012 | B2 |
| 8145379 | Schwinke | Mar 2012 | B2 |
| 8190322 | Lin et al. | May 2012 | B2 |
| 8209259 | Graham, Jr. et al. | Jun 2012 | B2 |
| 8212667 | Petite et al. | Jul 2012 | B2 |
| 8271473 | Berg | Sep 2012 | B2 |
| 8271547 | Taylor et al. | Sep 2012 | B2 |
| 8275717 | Ullman et al. | Sep 2012 | B2 |
| 8285439 | Hodges | Oct 2012 | B2 |
| 8296007 | Swaminathan et al. | Oct 2012 | B2 |
| 8311905 | Campbell et al. | Nov 2012 | B1 |
| 8355950 | Colson et al. | Jan 2013 | B2 |
| 8407664 | Moosmann et al. | Mar 2013 | B2 |
| 8428815 | Van Engelshoven et al. | Apr 2013 | B2 |
| 8438310 | Muilenburg et al. | May 2013 | B2 |
| 8448057 | Sugnet | May 2013 | B1 |
| 8521654 | Ford et al. | Aug 2013 | B2 |
| 8538894 | Ullman et al. | Sep 2013 | B2 |
| 8645193 | Swinson et al. | Feb 2014 | B2 |
| 8676638 | Blair et al. | Mar 2014 | B1 |
| 8725341 | Ogasawara | May 2014 | B2 |
| 8745641 | Coker | Jun 2014 | B1 |
| 8849689 | Jagannathan et al. | Sep 2014 | B1 |
| 8886389 | Edwards et al. | Nov 2014 | B2 |
| 8924071 | Stanek et al. | Dec 2014 | B2 |
| 8954222 | Costantino | Feb 2015 | B2 |
| 8996230 | Lorenz et al. | Mar 2015 | B2 |
| 8996235 | Singh et al. | Mar 2015 | B2 |
| 9014908 | Chen et al. | Apr 2015 | B2 |
| 9015059 | Sims et al. | Apr 2015 | B2 |
| 9026304 | Olsen, III et al. | May 2015 | B2 |
| 9047722 | Kurnik et al. | Jun 2015 | B2 |
| 9122716 | Naganathan et al. | Sep 2015 | B1 |
| 9165413 | Jones et al. | Oct 2015 | B2 |
| 9183681 | Fish | Nov 2015 | B2 |
| 9325650 | Yalavarty et al. | Apr 2016 | B2 |
| 9349223 | Palmer | May 2016 | B1 |
| 9384597 | Koch et al. | Jul 2016 | B2 |
| 9455969 | Cabrera et al. | Sep 2016 | B1 |
| 9477936 | Lawson et al. | Oct 2016 | B2 |
| 9577866 | Rogers et al. | Feb 2017 | B2 |
| 9596287 | Rybak et al. | Mar 2017 | B2 |
| 9619945 | Adderly et al. | Apr 2017 | B2 |
| 9659495 | Modica et al. | May 2017 | B2 |
| 9706008 | Rajan et al. | Jul 2017 | B2 |
| 9715665 | Schondorf et al. | Jul 2017 | B2 |
| 9754304 | Taira et al. | Sep 2017 | B2 |
| 9778045 | Bang | Oct 2017 | B2 |
| 9836714 | Lander et al. | Dec 2017 | B2 |
| 9983982 | Kumar et al. | Mar 2018 | B1 |
| 10032139 | Adderly et al. | Jul 2018 | B2 |
| 10083411 | Kinsey et al. | Sep 2018 | B2 |
| 10169607 | Sheth et al. | Jan 2019 | B1 |
| 10229394 | Davis et al. | Mar 2019 | B1 |
| 10448120 | Bursztyn et al. | Oct 2019 | B1 |
| 10475256 | Chowdhury et al. | Nov 2019 | B2 |
| 10509696 | Gilderman et al. | Dec 2019 | B1 |
| 10541938 | Timmerman et al. | Jan 2020 | B1 |
| 10552871 | Chadwick | Feb 2020 | B1 |
| 10657707 | Leise | May 2020 | B1 |
| 11080105 | Birkett et al. | Aug 2021 | B1 |
| 11117253 | Oleynik | Sep 2021 | B2 |
| 11190608 | Amar et al. | Nov 2021 | B2 |
| 11282041 | Sanderford et al. | Mar 2022 | B2 |
| 11322247 | Bullington et al. | May 2022 | B2 |
| 11392855 | Murakonda | Jul 2022 | B1 |
| 11443275 | Prakash et al. | Sep 2022 | B1 |
| 11468089 | Bales et al. | Oct 2022 | B1 |
| 11507892 | Henckel et al. | Nov 2022 | B1 |
| 20010005831 | Lewin et al. | Jun 2001 | A1 |
| 20010014868 | Herz et al. | Aug 2001 | A1 |
| 20010037332 | Miller et al. | Nov 2001 | A1 |
| 20010039594 | Park et al. | Nov 2001 | A1 |
| 20010054049 | Maeda et al. | Dec 2001 | A1 |
| 20020023111 | Arora et al. | Feb 2002 | A1 |
| 20020024537 | Jones et al. | Feb 2002 | A1 |
| 20020026359 | Long et al. | Feb 2002 | A1 |
| 20020032626 | Dewolf et al. | Mar 2002 | A1 |
| 20020032701 | Gao et al. | Mar 2002 | A1 |
| 20020042738 | Srinivasan et al. | Apr 2002 | A1 |
| 20020046245 | Hillar et al. | Apr 2002 | A1 |
| 20020049831 | Platner et al. | Apr 2002 | A1 |
| 20020052778 | Murphy et al. | May 2002 | A1 |
| 20020059260 | Jas | May 2002 | A1 |
| 20020065698 | Schick et al. | May 2002 | A1 |
| 20020065739 | Florance et al. | May 2002 | A1 |
| 20020069110 | Sonnenberg | Jun 2002 | A1 |
| 20020073080 | Lipkin | Jun 2002 | A1 |
| 20020082978 | Ghouri et al. | Jun 2002 | A1 |
| 20020091755 | Narin | Jul 2002 | A1 |
| 20020107739 | Schlee | Aug 2002 | A1 |
| 20020111727 | Vanstory et al. | Aug 2002 | A1 |
| 20020111844 | Vanstory et al. | Aug 2002 | A1 |
| 20020116197 | Erten | Aug 2002 | A1 |
| 20020116418 | Lachhwani et al. | Aug 2002 | A1 |
| 20020123359 | Wei et al. | Sep 2002 | A1 |
| 20020124053 | Adams et al. | Sep 2002 | A1 |
| 20020128728 | Murakami et al. | Sep 2002 | A1 |
| 20020129054 | Ferguson et al. | Sep 2002 | A1 |
| 20020133273 | Lowrey et al. | Sep 2002 | A1 |
| 20020138331 | Hosea et al. | Sep 2002 | A1 |
| 20020143646 | Boyden et al. | Oct 2002 | A1 |
| 20020154146 | Rodriquez et al. | Oct 2002 | A1 |
| 20020169851 | Weathersby et al. | Nov 2002 | A1 |
| 20020173885 | Lowrey et al. | Nov 2002 | A1 |
| 20020188869 | Patrick | Dec 2002 | A1 |
| 20020196273 | Krause | Dec 2002 | A1 |
| 20020198761 | Ryan et al. | Dec 2002 | A1 |
| 20020198878 | Baxter et al. | Dec 2002 | A1 |
| 20030014443 | Bernstein et al. | Jan 2003 | A1 |
| 20030023632 | Ries et al. | Jan 2003 | A1 |
| 20030033378 | Needham et al. | Feb 2003 | A1 |
| 20030036832 | Kokes et al. | Feb 2003 | A1 |
| 20030036964 | Boyden et al. | Feb 2003 | A1 |
| 20030037263 | Kamat et al. | Feb 2003 | A1 |
| 20030046179 | Anabtawi et al. | Mar 2003 | A1 |
| 20030051022 | Sogabe et al. | Mar 2003 | A1 |
| 20030055666 | Roddy et al. | Mar 2003 | A1 |
| 20030061263 | Riddle | Mar 2003 | A1 |
| 20030065532 | Takaoka | Apr 2003 | A1 |
| 20030065583 | Takaoka | Apr 2003 | A1 |
| 20030069785 | Lohse | Apr 2003 | A1 |
| 20030069790 | Kane | Apr 2003 | A1 |
| 20030074392 | Campbell et al. | Apr 2003 | A1 |
| 20030095038 | Dix | May 2003 | A1 |
| 20030101262 | Godwin | May 2003 | A1 |
| 20030115292 | Griffin et al. | Jun 2003 | A1 |
| 20030120502 | Robb et al. | Jun 2003 | A1 |
| 20030145310 | Thames et al. | Jul 2003 | A1 |
| 20030177050 | Crampton et al. | Sep 2003 | A1 |
| 20030177175 | Worley et al. | Sep 2003 | A1 |
| 20030225853 | Wang et al. | Dec 2003 | A1 |
| 20030229623 | Chang et al. | Dec 2003 | A1 |
| 20030233246 | Snapp et al. | Dec 2003 | A1 |
| 20040012631 | Skorski | Jan 2004 | A1 |
| 20040039646 | Hacker | Feb 2004 | A1 |
| 20040041818 | White et al. | Mar 2004 | A1 |
| 20040073546 | Forster et al. | Apr 2004 | A1 |
| 20040073564 | Haber et al. | Apr 2004 | A1 |
| 20040088228 | Mercer et al. | May 2004 | A1 |
| 20040093243 | Bodin et al. | May 2004 | A1 |
| 20040117046 | Colle et al. | Jun 2004 | A1 |
| 20040122735 | Meshkin et al. | Jun 2004 | A1 |
| 20040128320 | Grove et al. | Jul 2004 | A1 |
| 20040139203 | Graham, Jr. et al. | Jul 2004 | A1 |
| 20040148342 | Cotte | Jul 2004 | A1 |
| 20040156020 | Edwards | Aug 2004 | A1 |
| 20040163047 | Nagahara et al. | Aug 2004 | A1 |
| 20040181464 | Vanker et al. | Sep 2004 | A1 |
| 20040199413 | Hauser et al. | Oct 2004 | A1 |
| 20040220863 | Porter et al. | Nov 2004 | A1 |
| 20040225664 | Casement | Nov 2004 | A1 |
| 20040230897 | Latzel | Nov 2004 | A1 |
| 20040255233 | Croney et al. | Dec 2004 | A1 |
| 20040267263 | May | Dec 2004 | A1 |
| 20040268225 | Walsh et al. | Dec 2004 | A1 |
| 20040268232 | Tunning | Dec 2004 | A1 |
| 20050015491 | Koeppel | Jan 2005 | A1 |
| 20050021197 | Zimmerman et al. | Jan 2005 | A1 |
| 20050027611 | Wharton | Feb 2005 | A1 |
| 20050043614 | Huizenga | Feb 2005 | A1 |
| 20050065804 | Worsham et al. | Mar 2005 | A1 |
| 20050096963 | Myr et al. | May 2005 | A1 |
| 20050108112 | Ellenson et al. | May 2005 | A1 |
| 20050114270 | Hind et al. | May 2005 | A1 |
| 20050114764 | Gudenkauf et al. | May 2005 | A1 |
| 20050108637 | Sahota et al. | Jun 2005 | A1 |
| 20050149398 | McKay | Jul 2005 | A1 |
| 20050171836 | Leacy | Aug 2005 | A1 |
| 20050176482 | Raisinghani et al. | Aug 2005 | A1 |
| 20050187834 | Painter et al. | Aug 2005 | A1 |
| 20050198121 | Daniels et al. | Sep 2005 | A1 |
| 20050228736 | Norman et al. | Oct 2005 | A1 |
| 20050256755 | Chand et al. | Nov 2005 | A1 |
| 20050267774 | Merritt et al. | Dec 2005 | A1 |
| 20050268282 | Laird | Dec 2005 | A1 |
| 20050289020 | Bruns et al. | Dec 2005 | A1 |
| 20050289599 | Matsuura et al. | Dec 2005 | A1 |
| 20060004725 | Abraido-Fandino | Jan 2006 | A1 |
| 20060031811 | Ernst et al. | Feb 2006 | A1 |
| 20060059253 | Goodman et al. | Mar 2006 | A1 |
| 20060064637 | Rechterman et al. | Mar 2006 | A1 |
| 20060123330 | Horiuchi et al. | Jun 2006 | A1 |
| 20060129423 | Sheinson et al. | Jun 2006 | A1 |
| 20060129982 | Doyle | Jun 2006 | A1 |
| 20060136105 | Larson | Jun 2006 | A1 |
| 20060161841 | Horiuchi et al. | Jul 2006 | A1 |
| 20060200751 | Underwood et al. | Sep 2006 | A1 |
| 20060224447 | Koningstein | Oct 2006 | A1 |
| 20060248205 | Randle et al. | Nov 2006 | A1 |
| 20060248442 | Rosenstein et al. | Nov 2006 | A1 |
| 20060265355 | Taylor | Nov 2006 | A1 |
| 20060271844 | Suklikar | Nov 2006 | A1 |
| 20060277588 | Harrington et al. | Dec 2006 | A1 |
| 20060282328 | Gerace et al. | Dec 2006 | A1 |
| 20060282547 | Hasha et al. | Dec 2006 | A1 |
| 20070005446 | Fusz et al. | Jan 2007 | A1 |
| 20070016486 | Stone et al. | Jan 2007 | A1 |
| 20070027754 | Collins et al. | Feb 2007 | A1 |
| 20070033087 | Combs et al. | Feb 2007 | A1 |
| 20070033520 | Kimzey et al. | Feb 2007 | A1 |
| 20070053513 | Hoffberg | Mar 2007 | A1 |
| 20070100519 | Engel | May 2007 | A1 |
| 20070150368 | Arora et al. | Jun 2007 | A1 |
| 20070209011 | Padmanabhuni et al. | Sep 2007 | A1 |
| 20070226540 | Konieczny | Sep 2007 | A1 |
| 20070250229 | Wu | Oct 2007 | A1 |
| 20070250327 | Hedy | Oct 2007 | A1 |
| 20070250840 | Coker et al. | Oct 2007 | A1 |
| 20070271154 | Broudy et al. | Nov 2007 | A1 |
| 20070271330 | Mattox et al. | Nov 2007 | A1 |
| 20070271389 | Joshi et al. | Nov 2007 | A1 |
| 20070282711 | Ullman et al. | Dec 2007 | A1 |
| 20070282712 | Ullman et al. | Dec 2007 | A1 |
| 20070282713 | Ullman et al. | Dec 2007 | A1 |
| 20070288413 | Mizuno et al. | Dec 2007 | A1 |
| 20070294192 | Tellefsen | Dec 2007 | A1 |
| 20070299940 | Gbadegesin et al. | Dec 2007 | A1 |
| 20080010561 | Bay et al. | Jan 2008 | A1 |
| 20080015921 | Libman | Jan 2008 | A1 |
| 20080015929 | Koeppel et al. | Jan 2008 | A1 |
| 20080027827 | Eglen et al. | Jan 2008 | A1 |
| 20080119983 | Inbarajan et al. | May 2008 | A1 |
| 20080172632 | Stambaugh | Jul 2008 | A1 |
| 20080189143 | Wurster | Aug 2008 | A1 |
| 20080195435 | Bentley et al. | Aug 2008 | A1 |
| 20080195932 | Oikawa et al. | Aug 2008 | A1 |
| 20080201163 | Barker et al. | Aug 2008 | A1 |
| 20080255925 | Vailaya et al. | Oct 2008 | A1 |
| 20090012887 | Taub et al. | Jan 2009 | A1 |
| 20090024918 | Silverbrook et al. | Jan 2009 | A1 |
| 20090043780 | Hentrich, Jr. et al. | Feb 2009 | A1 |
| 20090070435 | Abhyanker | Mar 2009 | A1 |
| 20090089134 | Uyeki | Apr 2009 | A1 |
| 20090106036 | Tamura et al. | Apr 2009 | A1 |
| 20090112687 | Blair et al. | Apr 2009 | A1 |
| 20090138329 | Wanker | May 2009 | A1 |
| 20090182232 | Zhang et al. | Jul 2009 | A1 |
| 20090187513 | Noy et al. | Jul 2009 | A1 |
| 20090187939 | Lajoie | Jul 2009 | A1 |
| 20090198507 | Rhodus | Aug 2009 | A1 |
| 20090204454 | Lagudi | Aug 2009 | A1 |
| 20090204655 | Wendelberger | Aug 2009 | A1 |
| 20090222532 | Finlaw | Sep 2009 | A1 |
| 20090265607 | Raz et al. | Oct 2009 | A1 |
| 20090313035 | Esser et al. | Dec 2009 | A1 |
| 20100011415 | Cortes et al. | Jan 2010 | A1 |
| 20100023393 | Costy et al. | Jan 2010 | A1 |
| 20100070343 | Taira et al. | Mar 2010 | A1 |
| 20100082778 | Muilenburg et al. | Apr 2010 | A1 |
| 20100082780 | Muilenburg et al. | Apr 2010 | A1 |
| 20100088158 | Pollack | Apr 2010 | A1 |
| 20100100259 | Geiter | Apr 2010 | A1 |
| 20100100506 | Marot | Apr 2010 | A1 |
| 20100131363 | Sievert et al. | May 2010 | A1 |
| 20100235219 | Merrick et al. | Sep 2010 | A1 |
| 20100235231 | Jewer | Sep 2010 | A1 |
| 20100293030 | Wu | Nov 2010 | A1 |
| 20100312608 | Shan et al. | Dec 2010 | A1 |
| 20100318408 | Sankaran et al. | Dec 2010 | A1 |
| 20100324777 | Tominaga et al. | Dec 2010 | A1 |
| 20110010432 | Uyeki | Jan 2011 | A1 |
| 20110015989 | Tidwell et al. | Jan 2011 | A1 |
| 20110022525 | Swinson et al. | Jan 2011 | A1 |
| 20110082804 | Swinson et al. | Apr 2011 | A1 |
| 20110145064 | Anderson et al. | Jun 2011 | A1 |
| 20110161167 | Jallapuram | Jun 2011 | A1 |
| 20110191264 | Inghelbrecht et al. | Aug 2011 | A1 |
| 20110196762 | Dupont | Aug 2011 | A1 |
| 20110224864 | Gellatly et al. | Sep 2011 | A1 |
| 20110231055 | Knight et al. | Sep 2011 | A1 |
| 20110288937 | Manoogian, III | Nov 2011 | A1 |
| 20110307296 | Hall et al. | Dec 2011 | A1 |
| 20110307411 | Bolivar et al. | Dec 2011 | A1 |
| 20120066010 | Williams et al. | Mar 2012 | A1 |
| 20120089474 | Xiao et al. | Apr 2012 | A1 |
| 20120095804 | Calabrese et al. | Apr 2012 | A1 |
| 20120116868 | Chin et al. | May 2012 | A1 |
| 20120158211 | Chen et al. | Jun 2012 | A1 |
| 20120209714 | Douglas et al. | Aug 2012 | A1 |
| 20120221125 | Bell | Aug 2012 | A1 |
| 20120265648 | Jerome et al. | Oct 2012 | A1 |
| 20120268294 | Michaelis et al. | Oct 2012 | A1 |
| 20120278886 | Luna | Nov 2012 | A1 |
| 20120284113 | Pollak | Nov 2012 | A1 |
| 20120316981 | Hoover et al. | Dec 2012 | A1 |
| 20130046432 | Edwards et al. | Feb 2013 | A1 |
| 20130080196 | Schroeder et al. | Mar 2013 | A1 |
| 20130080305 | Virag et al. | Mar 2013 | A1 |
| 20130151334 | Berkhin et al. | Jun 2013 | A1 |
| 20130151468 | Wu et al. | Jun 2013 | A1 |
| 20130191445 | Gayman et al. | Jul 2013 | A1 |
| 20130204484 | Ricci | Aug 2013 | A1 |
| 20130226699 | Long | Aug 2013 | A1 |
| 20130317864 | Tofte | Nov 2013 | A1 |
| 20130325541 | Capriotti et al. | Dec 2013 | A1 |
| 20130332023 | Bertosa et al. | Dec 2013 | A1 |
| 20140012659 | Yan | Jan 2014 | A1 |
| 20140026037 | Garb et al. | Jan 2014 | A1 |
| 20140052327 | Hosein et al. | Feb 2014 | A1 |
| 20140081675 | Ives | Mar 2014 | A1 |
| 20140088866 | Knapp et al. | Mar 2014 | A1 |
| 20140094992 | Lambert et al. | Apr 2014 | A1 |
| 20140122178 | Knight | May 2014 | A1 |
| 20140136278 | Carvalho | May 2014 | A1 |
| 20140229207 | Swamy | Aug 2014 | A1 |
| 20140229391 | East et al. | Aug 2014 | A1 |
| 20140244110 | Tharaldson et al. | Aug 2014 | A1 |
| 20140277906 | Lowrey et al. | Sep 2014 | A1 |
| 20140278805 | Thompson | Sep 2014 | A1 |
| 20140316825 | Van Dijk et al. | Oct 2014 | A1 |
| 20140324275 | Stanek et al. | Oct 2014 | A1 |
| 20140324536 | Cotton | Oct 2014 | A1 |
| 20140331301 | Subramani et al. | Nov 2014 | A1 |
| 20140337163 | Whisnant | Nov 2014 | A1 |
| 20140337825 | Challa et al. | Nov 2014 | A1 |
| 20140379530 | Kim et al. | Dec 2014 | A1 |
| 20140379817 | Logue et al. | Dec 2014 | A1 |
| 20150032546 | Calman et al. | Jan 2015 | A1 |
| 20150057875 | McGinnis et al. | Feb 2015 | A1 |
| 20150058151 | Sims et al. | Feb 2015 | A1 |
| 20150066781 | Johnson et al. | Mar 2015 | A1 |
| 20150066933 | Kolodziej et al. | Mar 2015 | A1 |
| 20150100199 | Kurnik et al. | Apr 2015 | A1 |
| 20150142256 | Jones | May 2015 | A1 |
| 20150142535 | Payne et al. | May 2015 | A1 |
| 20150207701 | Faaborg et al. | Jul 2015 | A1 |
| 20150227894 | Mapes, Jr. et al. | Aug 2015 | A1 |
| 20150242819 | Moses et al. | Aug 2015 | A1 |
| 20150248761 | Dong | Sep 2015 | A1 |
| 20150254591 | Raskind | Sep 2015 | A1 |
| 20150268059 | Borghesani et al. | Sep 2015 | A1 |
| 20150268975 | Du et al. | Sep 2015 | A1 |
| 20150278886 | Fusz | Oct 2015 | A1 |
| 20150286475 | Vangelov et al. | Oct 2015 | A1 |
| 20150286979 | Ming et al. | Oct 2015 | A1 |
| 20150290795 | Oleynik | Oct 2015 | A1 |
| 20150334165 | Arling et al. | Nov 2015 | A1 |
| 20160004516 | Ivanov et al. | Jan 2016 | A1 |
| 20160059412 | Oleynik | Mar 2016 | A1 |
| 20160071054 | Kakarala et al. | Mar 2016 | A1 |
| 20160092944 | Taylor et al. | Mar 2016 | A1 |
| 20160132935 | Shen et al. | May 2016 | A1 |
| 20160140609 | Demir | May 2016 | A1 |
| 20160140620 | Pinkowish et al. | May 2016 | A1 |
| 20160140622 | Wang et al. | May 2016 | A1 |
| 20160148439 | Akselrod et al. | May 2016 | A1 |
| 20160162817 | Grimaldi et al. | Jun 2016 | A1 |
| 20160179968 | Ormseth et al. | Jun 2016 | A1 |
| 20160180358 | Battista | Jun 2016 | A1 |
| 20160180378 | Toshida et al. | Jun 2016 | A1 |
| 20160180418 | Jaeger | Jun 2016 | A1 |
| 20160267503 | Zakai-Or et al. | Sep 2016 | A1 |
| 20160275533 | Smith et al. | Sep 2016 | A1 |
| 20160277510 | Du et al. | Sep 2016 | A1 |
| 20160307174 | Marcelle et al. | Oct 2016 | A1 |
| 20160335727 | Jimenez | Nov 2016 | A1 |
| 20160337278 | Peruri et al. | Nov 2016 | A1 |
| 20160357599 | Glatfelter | Dec 2016 | A1 |
| 20160371641 | Wilson et al. | Dec 2016 | A1 |
| 20170034547 | Jain et al. | Feb 2017 | A1 |
| 20170039785 | Richter et al. | Feb 2017 | A1 |
| 20170053460 | Hauser et al. | Feb 2017 | A1 |
| 20170060929 | Chesla et al. | Mar 2017 | A1 |
| 20170064038 | Chen | Mar 2017 | A1 |
| 20170093700 | Gilley et al. | Mar 2017 | A1 |
| 20170124525 | Johnson et al. | May 2017 | A1 |
| 20170126848 | George et al. | May 2017 | A1 |
| 20170206465 | Jin | Jul 2017 | A1 |
| 20170262894 | Kirti et al. | Sep 2017 | A1 |
| 20170293894 | Taliwal et al. | Oct 2017 | A1 |
| 20170308844 | Kelley | Oct 2017 | A1 |
| 20170308864 | Kelley | Oct 2017 | A1 |
| 20170308865 | Kelley | Oct 2017 | A1 |
| 20170316459 | Strauss et al. | Nov 2017 | A1 |
| 20170337573 | Toprak | Nov 2017 | A1 |
| 20170352054 | Ma et al. | Dec 2017 | A1 |
| 20170359216 | Naiden et al. | Dec 2017 | A1 |
| 20170364733 | Estrada | Dec 2017 | A1 |
| 20180067932 | Paterson et al. | Mar 2018 | A1 |
| 20180074864 | Chen et al. | Mar 2018 | A1 |
| 20180095733 | Torman et al. | Apr 2018 | A1 |
| 20180173806 | Forstmann et al. | Jun 2018 | A1 |
| 20180204281 | Painter et al. | Jul 2018 | A1 |
| 20180225710 | Kar et al. | Aug 2018 | A1 |
| 20180232749 | Moore, Jr. et al. | Aug 2018 | A1 |
| 20180285901 | Zackrone | Oct 2018 | A1 |
| 20180285925 | Zackrone | Oct 2018 | A1 |
| 20180300124 | Malladi et al. | Oct 2018 | A1 |
| 20190028360 | Douglas et al. | Jan 2019 | A1 |
| 20190073641 | Utke | Mar 2019 | A1 |
| 20190114330 | Xu et al. | Apr 2019 | A1 |
| 20190213426 | Chen | Jul 2019 | A1 |
| 20190294878 | Endras | Sep 2019 | A1 |
| 20190297162 | Amar et al. | Sep 2019 | A1 |
| 20190334884 | Ross et al. | Oct 2019 | A1 |
| 20200019388 | Jaeger et al. | Jan 2020 | A1 |
| 20200038363 | Kim | Feb 2020 | A1 |
| 20200050879 | Zaman | Feb 2020 | A1 |
| 20200066067 | Herman | Feb 2020 | A1 |
| 20200118365 | Wang | Apr 2020 | A1 |
| 20200177476 | Agarwal et al. | Jun 2020 | A1 |
| 20200327371 | Sharma et al. | Oct 2020 | A1 |
| 20210072976 | Chintagunta et al. | Mar 2021 | A1 |
| 20210090694 | Colley et al. | Mar 2021 | A1 |
| 20210157562 | Sethi et al. | May 2021 | A1 |
| 20210184780 | Yang et al. | Jun 2021 | A1 |
| 20210224975 | Ranca | Jul 2021 | A1 |
| 20210240657 | Kumar et al. | Aug 2021 | A1 |
| 20210256616 | Hayward | Aug 2021 | A1 |
| 20210287106 | Jerram | Sep 2021 | A1 |
| 20210303644 | Shear | Sep 2021 | A1 |
| 20210350334 | Ave et al. | Nov 2021 | A1 |
| 20210359940 | Shen et al. | Nov 2021 | A1 |
| 20220020086 | Kuchenbecker et al. | Jan 2022 | A1 |
| 20220028928 | Seo et al. | Jan 2022 | A1 |
| 20220046105 | Amar et al. | Feb 2022 | A1 |
| 20220172723 | Tendolkar et al. | Jun 2022 | A1 |
| 20220191663 | Karpoor et al. | Jun 2022 | A1 |
| 20220208319 | Ansari et al. | Jun 2022 | A1 |
| 20220237084 | Bhagi et al. | Jul 2022 | A1 |
| 20220237171 | Bailey et al. | Jul 2022 | A1 |
| 20220293107 | Leaman et al. | Sep 2022 | A1 |
| 20220300735 | Kelly et al. | Sep 2022 | A1 |
| 20230214892 | Christian et al. | Jul 2023 | A1 |
| Number | Date | Country |
|---|---|---|
| 2494350 | May 2004 | CA |
| 0461888 | Mar 1995 | EP |
| 2007002759 | Jan 2007 | WO |
| Entry |
|---|
| http://web.archive.org/web/20010718130244/http://chromedata.com/maing2/about/index.asp, 1 pg. |
| http://web.archive.org/web/20050305055408/http://www.dealerclick.com/, 1 pg. |
| http://web.archive.org/web/20050528073821/http://www.kbb.com/, 1 pg. |
| http://web.archive.org/web/20050531000823/http://www.carfax.com/, 1 pg. |
| Internet Archive Dan Gillmor Sep. 1, 1996. |
| Internet Archive Wayback Machine, archive of LDAP Browser.com—FAQ. Archived Dec. 11, 2000. Available at <http://web.archive.org/web/200012110152/http://www.ldapbrowser.com/faq/faq.php3?sID=fe4ae66f023d86909f35e974f3a1ce>. |
| Internet Archive Wayback Machine, archive of LDAP Browser.com—Product Info. Archived Dec. 11, 2000. Available at <http://web.archive.org/web/200012110541/http://www.ldapbrowser.com/prodinfo/prodinfo.php3?sID=fe4ae66f2fo23d86909f35e974f3a1ce>. |
| Internet Archive: Audio Archive, http://www.archive.org/audio/audio-searchresults.php?search=@start=0&limit=100&sort=ad, printed May 12, 2004, 12 pgs. |
| Internet Archive: Democracy Now, http://www.archive.org/audio/collection.php?collection=democracy_now, printed May 12, 2004, 2 pgs. |
| Java 2 Platform, Enterprise Edition (J2EE) Overview, printed Mar. 6, 2010, 3 pgs. |
| Java version history—Wikipedia, the free encyclopedia, printed Mar. 6, 2010, 9 pgs. |
| Permissions in the Java™ 2 SDK, printed Mar. 6, 2010, 45 pgs. |
| Trademark Application, Serial No. 76375405. 13 pages of advertising material and other application papers enclosed. Available from Trademark Document Retrieval system at. |
| Trademark Electronic Search System record for Serial No. 76375405, Word Mark “NITRA”. |
| “An Appointment with Destiny—The Time for Web-Enabled Scheduling has Arrived”, Link Fall, 2007, 2 pages. |
| “How a Solution found a Problem of Scheduling Service Appointments”, Automotive News, 2016, 4 pages. |
| “IBM Tivoli Access Manager Base Administration Guide”, Version 5.1. International Business Machines Corporation. Entire book enclosed and cited., 2003, 402 pgs. |
| “NetFormx Offers Advanced Network Discovery Software”, PR Newswire. Retrieved from http://www.highbeam.com/doc/1G1-54102907.html>., Mar. 15, 1999. |
| “Openbay Announces First-of-its-Kind Connected Car Repair Service”, openbay.com, Mar. 31, 2015, 14 pages. |
| “Service Advisor”, Automotive Dealership Institute, 2007, 26 pages. |
| “xTime.com Web Pages”, Jan. 8, 2015, 1 page. |
| “xTimes Newsletter”, vol. 7, 2013, 4 pages. |
| U.S. Appl. No. 10/350,795 , Non-Final Office Action, Dec. 26, 2008, 13 pages. |
| U.S. Appl. No. 10/350,795 , Non-Final Office Action, Feb. 6, 2006, 11 pages. |
| U.S. Appl. No. 10/350,795 , Non-Final Office Action, Jul. 22, 2009, 22 pages. |
| U.S. Appl. No. 10/350,795 , Final Office Action, Jul. 6, 2041, 26 pages. |
| U.S. Appl. No. 10/350,795 , Non-Final Office Action, Jun. 29, 2006, 11 pages. |
| U.S. Appl. No. 10/350,795 , Non-Final Office Action, Mar. 12, 2007, 10 pages. |
| U.S. Appl. No. 10/350,795 , Final Office Action, Mar. 3, 2010, 24 pages. |
| U.S. Appl. No. 10/350,795 , Non-Final Office Action, May 29, 2008, 10 pages. |
| U.S. Appl. No. 10/350,795 , Notice of Allowance, May 7, 2012, 15 pages. |
| U.S. Appl. No. 10/350,795 , Non-Final Office Action, Nov. 1, 2010, 19 pages. |
| U.S. Appl. No. 10/350,796 , Notice of Allowance, Feb. 1, 2006, 5 pages. |
| U.S. Appl. No. 10/350,796 , Non-Final Office Action, May 19, 2005, 7 pages. |
| U.S. Appl. No. 10/350,810 , Notice of Allowance, Apr. 14, 2008, 6 pages. |
| U.S. Appl. No. 10/350,810 , Non-Final Office Action, Apr. 17, 2007, 12 pages. |
| U.S. Appl. No. 10/350,810 , Final Office Action, Apr. 5, 2005, 12 pages. |
| U.S. Appl. No. 10/350,810 , Notice of Non-compliant Amendment, Dec. 12, 2006. |
| U.S. Appl. No. 10/350,810 , Non-Final Office Action, Dec. 9, 2005, 14 pages. |
| U.S. Appl. No. 10/350,810 , Final Office Action, May 18, 2006, 15 pages. |
| U.S. Appl. No. 10/350,810 , Final Office Action, Nov. 14, 2007, 13 pages. |
| U.S. Appl. No. 10/350,810 , Non-Final Office Action, Sep. 22, 2004, 10 pages. |
| U.S. Appl. No. 10/351,465 , Non-Final Office Action, Jul. 27, 2004, 9 pages. |
| U.S. Appl. No. 10/351,465 , Final Office Action, May 5, 2005, 8 pages. |
| U.S. Appl. No. 10/351,465 , Notice of Allowance, Sep. 21, 2005, 4 pages. |
| U.S. Appl. No. 10/351,606 , Notice of Allowance, Apr. 4, 2006, 12 pages. |
| U.S. Appl. No. 10/351,606 , Non-final Office Action, May 17, 2004, 5 pages. |
| U.S. Appl. No. 10/351,606 , Non-final Office Action, Dec. 19, 2005, 7 pages. |
| U.S. Appl. No. 10/655,899, Non-Final Office Action, Sep. 17, 2007, 11 pages. |
| U.S. Appl. No. 10/665,899 , Non-Final Office Action, Aug. 30, 2010, 23 pages. |
| U.S. Appl. No. 16/951,833 , Notice of Allowance, Jun. 16, 2021, 14 pages. |
| Hu, Bo , “A Platform based Distributed Service Framework for Large-scale Cloud Ecosystem Development”, IEEE Computer Society, 2015, 8 pages. |
| U.S. Appl. No. 15/478,042, Non-Final Office Action, Nov. 19, 2021, 45 pages. |
| U.S. Appl. No. 16/911,154 , Final Office Action, Mar. 28, 2022, 17 pages. |
| U.S. Appl. No. 17/156,254, Non-Final Office Action, Feb. 25, 2022, 18 pages. |
| U.S. Appl. No. 13/025,019 , Non-Final Office Action, Oct. 6, 2017. |
| U.S. Appl. No. 13/025,019 , Final Office Action, Sep. 12, 2013, 13 pages. |
| U.S. Appl. No. 13/025,019 , Non-Final Office Action, Sep. 18, 2014, 15 pages. |
| U.S. Appl. No. 13/025,019 , Notice of Allowance, Sep. 26, 2019, 9 pages. |
| U.S. Appl. No. 14/208,042 , Final Office Action, Apr. 16, 2018. |
| U.S. Appl. No. 14/208,042 , Non-Final Office Action, Aug. 21, 2020, 13 pages. |
| U.S. Appl. No. 14/208,042 , Final Office Action, Dec. 6, 2016, 26 pages. |
| Chen, Deren “Business to Business Standard and Supply Chain System Framework in Virtual Enterprises”, Compuer Supported Cooperative Work in Design, The Sixth International Conference on 2001, pp. 472-476. |
| U.S. Appl. No. 14/208,042 , Final Office Action, Jan. 11, 2019, 16 pages. |
| U.S. Appl. No. 14/208,042 , Advisory Action, Jul. 12, 2018. |
| U.S. Appl. No. 14/208,042 , Non-Final Office Action, Jun. 30, 2016, 23 pages. |
| U.S. Appl. No. 14/208,042 , Notice of Allowance, May 6, 2021, 13 pages. |
| U.S. Appl. No. 14/208,042 , Non-Final Office Action, Sep. 20, 2017. |
| U.S. Appl. No. 14/208,042 , Non-Final Office Action, Sep. 21, 2018. |
| U.S. Appl. No. 15/134,779 , Final Office Action, Feb. 27, 2020, 18 pages. |
| U.S. Appl. No. 15/134,779 , Non-Final Office Action, Jan. 30, 2019, 26 pages. |
| U.S. Appl. No. 15/134,779 , Advisory Action, Jul. 29, 2019, 6 pages. |
| U.S. Appl. No. 15/134,779 , Final Office Action, May 17, 2019, 25 pages. |
| U.S. Appl. No. 15/134,779 , Non-Final Office Action, Nov. 19, 2019, 27 pages. |
| U.S. Appl. No. 15/134,779 , Notice of Allowance, Sep. 9, 2020, 12 pages. |
| U.S. Appl. No. 15/134,793 , Non-Final Office Action, Jan. 30, 2019, 26 pages. |
| U.S. Appl. No. 15/134,793 , Advisory Action, Jul. 29, 2019, 6 pages. |
| U.S. Appl. No. 15/134,793 , Final Office Action, Mar. 27, 2020, 22 pages. |
| U.S. Appl. No. 15/134,793 , Final Office Action, May 13, 2019, 26 pages. |
| U.S. Appl. No. 15/134,793 , Non-Final Office Action, Nov. 19, 2019, 31 pages. |
| U.S. Appl. No. 15/134,793 , Notice of Allowance, Nov. 2, 2020, 13 pages. |
| U.S. Appl. No. 15/134,820 , Non-Final Office Action, Feb. 23, 2018. |
| U.S. Appl. No. 15/134,820 , Notice of Allowance, Jan. 28, 2019, 7 pages. |
| U.S. Appl. No. 15/134,820 , Final Office Action, Sep. 21, 2018. |
| U.S. Appl. No. 15/478,042 , Non-Final Office Action, Aug. 4, 2020, 42 pages. |
| U.S. Appl. No. 15/478,042 , Final Office Action, Mar. 19, 2020, 35 pages. |
| U.S. Appl. No. 15/478,042 , Final Office Action, May 5, 2021, 38 pages. |
| U.S. Appl. No. 15/478,042 , Non-Final Office Action, Oct. 10, 2019, 26 pages. |
| U.S. Appl. No. 15/478,048 , Final Office Action, Apr. 9, 2020, 42 pages. |
| U.S. Appl. No. 15/478,048 , Non-Final Office Action, Mar. 8, 2021, 69 pages. |
| U.S. Appl. No. 15/478,048 , Non-Final Office Action, Sep. 30, 2019, 30 pages. |
| U.S. Appl. No. 15/602,999 , Notice of Allowance, Apr. 18, 2019, 6 pages. |
| U.S. Appl. No. 15/602,999 , Advisory Action, Jan. 31, 2019, 3 pages. |
| U.S. Appl. No. 15/602,999 , Non-Final Office Action, May 3, 2018. |
| U.S. Appl. No. 15/602,999 , Final Office Action, Nov. 21, 2018. |
| U.S. Appl. No. 16/041,552 , Final Office Action, Apr. 27, 2021, 23 pages. |
| U.S. Appl. No. 16/041,552 , Non-Final Office Action, Dec. 27, 2019, 13 pages. |
| U.S. Appl. No. 16/041,552 , Final Office Action, May 29, 2020, 18 pages. |
| U.S. Appl. No. 16/041,552 , Non-Final Office Action, Sep. 17, 2020, 16 pages. |
| U.S. Appl. No. 16/951,833 , Non-Final Office Action, Feb. 4, 2021, 10 pages. |
| Aloisio, Giovanni , et al., “Web-based access to the Grid using the Grid Resource Broker portal”, Google, 2002, pp. 1145-1160. |
| Anonymous , “Software ready for prime time”, Automotive News. Detroit, vol. 76, Issue 5996, Nov. 5, 2001, p. 28. |
| Bedell, Doug , Dallas Morning News, “I Know Someone Who Knows Kevin Bacon”. Oct. 27, 1998. 4 pgs. |
| Chadwick, D.W. , “Understanding X.500—The Directory”, Available at <http://sec.cs.kent.ac.uk/x500book/>. Entire work cited., 1996. |
| Chatterjee, Pallab , et al., “On-board diagnostics not just for racing anymore”, EDN.com, May 6, 2013, 7 pages. |
| Clemens Grelck , “A Multithread Compiler Backend for High-Level Array Programming”, 2003. |
| CNY Business Journal , “Frank La Voila named Southern Tier Small-Business Person of 1999”, Jun. 11, 1999, 2 pages. |
| Croswell, Wayne , “Service Shop Optimiztion”, Modern Tire Retailer, May 21, 2013, 7 pages. |
| Davis, Peter T., et al., “Sams Teach Yourself Microsoft Windows NT Server 4 in 21 Days”, Sams® Publishing, ISBN: 0-672-31555-6, 1999, printed Dec. 21, 2008, 15 pages. |
| Derfler, Frank J., et al., “How Networks Work: Millennium Edition”, Que, A Division of Macmillan Computer Publishing, ISBN: 0-7897-2445-6, 2000, 9 pages. |
| Drawbaugh, Ben , “Automatic Link Review: an expensive way to learn better driving habits”, Endgadget.com, Nov. 26, 2013, 14 pages. |
| Emmanuel, Daniel , “Basics to Creating an Appointment System for Automotive Service Customers”, Automotiveservicemanagement.com, 2006, 9 pages. |
| Hogue , et al., “Thresher: Automating the Unwrapping of Semantic Content from the World Wide Web”, ACM, 2005, pp. 86-95. |
| Housel, Barron C., et al., “WebExpress: A client/intercept based system for optimizing Web browsing in a wireless environment”, Google, 1998, pp. 419-431. |
| Interconnection , In Roget's II The New Thesaurus. Boston, MA: Houghton Mifflin http://www.credoreference.com/entry/hmrogets/interconnection, 2003, Retrieved Jul. 16, 2009, 1 page. |
| Jenkins, Will , “Real-time vehicle performance monitoring with data intergrity”, A Thesis Submitted to the Faculty of Mississippi State University, Oct. 2006, 57 pages. |
| Johns, Pamela , et al., “Competitive intelligence in service marketing”, Marketing Intelligence & Planning, vol. 28, No. 5, 2010, pp. 551-570. |
| Lavrinc, Damon , “First Android-powered infotainment system coming to 2012 Saab 9-3”, Autoblog.com, Mar. 2, 2011, 8 pages. |
| Lee, Adam J., et al., “Searching for Open Windows and Unlocked Doors: Port Scanning in Large-Scale Commodity Clusters”, Cluster Computing and the Grid, 2005. IEEE International Symposium on vol. 1, 2005, pp. 146-151. |
| Michener, J.R. , et al., “Managing System and Active-Content Integrity”, Computer; vol. 33, Issue: 7, 2000, pp. 108-110. |
| Milic-Frayling, Natasa , et al., “SmartView: Enhanced Document Viewer for Mobile Devices”, Google, Nov. 15, 2002, 11 pages. |
| Needham, Charlie , “Google Now Taking Appointments for Auto Repair Shops”, Autoshopsolutions.com, Aug. 25, 2015, 6 pages. |
| Open Bank Project , https://www.openbankproject.com/, retrieved Nov. 23, 2020, 10 pages. |
| openbay.com Web Pages , Openbay.com, retrieved from archive.org May 14, 2019, Apr. 2015, 6 pages. |
| openbay.com Web Pages , Openbay.com, retrieved from archive.org on May 14, 2019, Feb. 2014, 2 pages. |
| openbay.com Web Pages , Openbay.com, retrieved from archive.org, May 14, 2019, Mar. 2015, 11 pages. |
| Phelan, Mark , “Smart phone app aims to automate car repairs”, Detroit Free Press Auto Critic, Mar. 31, 2015, 2 pages. |
| Pubnub Staff , “Streaming Vehicle Data in Realtime with Automatic (Pt 1)”, Pubnub.com, Aug. 17, 2015, 13 pages. |
| Standards for Technology in Auto , https://www.starstandard.org/, retrieved Nov. 23, 2020, 4 pages. |
| Strebe, Matthew , et al., MCSE: NT Server 4 Study Guide, Third Edition. Sybex Inc. Front matter, 2000, pp. 284-293, and 308-347. |
| Warren, Tamara , “This Device Determines What Ails Your Car and Finds a Repair Shop—Automatically”, CarAndDriver.com, Apr. 8, 2015, 7 pages. |
| You, Song , et al., “Overview of Remote Diagnosis and Maintenance for Automotive Systems”, 2005 SAE World Congress, Apr. 11-14, 2015, 10 pages. |
| U.S. Appl. No. 10/665,899 , Final Office Action, Feb. 24, 2010, 22 pages. |
| U.S. Appl. No. 10/665,899 , Final Office Action, Jul. 7, 2008, 11 pages. |
| U.S. Appl. No. 10/665,899 , Final Office Action, Mar. 8, 2011, 21 pages. |
| U.S. Appl. No. 10/665,899 , Final Office Action, May 11, 2009, 14 pages. |
| U.S. Appl. No. 10/665,899 , Non-Final Office Action, Nov. 13, 2008, 11 pages. |
| U.S. Appl. No. 10/665,899 , Non-Final Office Action, Sep. 14, 2009, 14 pages. |
| U.S. Appl. No. 11/149,909 , Final Office Action, Feb. 4, 2009, 14 pages. |
| U.S. Appl. No. 11/149,909 , Non-Final Office Action, May 13, 2008, 14 pages. |
| U.S. Appl. No. 11/149,909 , Non-Final Office Action, May 6, 2009, 6 pages. |
| U.S. Appl. No. 11/149,909 , Notice of Allowance, Sep. 16, 2009, 7 pages. |
| U.S. Appl. No. 11/414,939 , Non-Final Office Action, Jul. 19, 2010, 7 pages. |
| U.S. Appl. No. 11/414,939 , Non-Final Office Action, Mar. 9, 2010, 11 pages. |
| U.S. Appl. No. 11/414,939 , Notice of Allowance, Nov. 2, 2010. |
| U.S. Appl. No. 11/442,821 , Final Office Action, Apr. 7, 2009, 19 pages. |
| U.S. Appl. No. 11/442,821 , Notice of Allowance, Jul. 30, 2012, 6 pages. |
| U.S. Appl. No. 11/442,821 , Non-Final Office Action, Jun. 1, 2011, 23 pages. |
| U.S. Appl. No. 11/442,821 , Final Office Action, May 21, 2010, 28 pages. |
| U.S. Appl. No. 11/442,821 , Non-Final Office Action, Nov. 12, 2009, 19 pages. |
| U.S. Appl. No. 11/442,821 , Final Office Action, Nov. 29, 2011, 26 pages. |
| U.S. Appl. No. 11/442,821 , Non-Final Office Action, Sep. 3, 2008, 14 pages. |
| U.S. Appl. No. 11/446,011 , Notice of Allowance, Aug. 9, 2011, 10 pages. |
| U.S. Appl. No. 11/446,011 , Final Office Action, Jun. 8, 2010, 12 pages. |
| U.S. Appl. No. 11/446,011 , Non-Final Office Action, Mar. 1, 2011, 15 pages. |
| U.S. Appl. No. 11/446,011 , Non-Final Office Action, Nov. 27, 2009, 14 pages. |
| U.S. Appl. No. 11/524,602 , Notice of Allowance, Aug. 6, 2013, 22 pages. |
| U.S. Appl. No. 11/524,602 , Non-Final Office Action, Dec. 11, 2009, 20 pages. |
| U.S. Appl. No. 11/524,602 , Final Office Action, Jul. 27, 2010, 13 pages. |
| U.S. Appl. No. 11/524,602 , Final Office Action, Jun. 26, 2012, 11 pages. |
| U.S. Appl. No. 11/524,602 , Non-Final Office Action, Nov. 14, 2011, 19 pages. |
| U.S. Appl. No. 11/525,009 , Non-Final Office Action, Aug. 10, 2011, 18 pages. |
| U.S. Appl. No. 11/525,009 , Final Office Action, Aug. 3, 2010, 16 pages. |
| U.S. Appl. No. 11/525,009 , Non-Final Office Action, Dec. 16, 2009, 20 pages. |
| U.S. Appl. No. 11/525,009 , Notice of Allowance, Jul. 23, 2012, 19 pages. |
| U.S. Appl. No. 12/243,852 , Restriction Requirement, Dec. 7, 2010. |
| U.S. Appl. No. 12/243,852 , Notice of Allowance, Feb. 27, 2013, 6 pages. |
| U.S. Appl. No. 12/243,852 , Non-Final Office Action, Jan. 16, 2013, 5 pages. |
| U.S. Appl. No. 12/243,852 , Non-Final Office Action, Mar. 17, 2011, 8 pages. |
| U.S. Appl. No. 12/243,852 , Supplemental Notice of Allowability, Mar. 19, 2013, 3 pages. |
| U.S. Appl. No. 12/243,852 , Final Office Action, Oct. 24, 2011, 13 pages. |
| U.S. Appl. No. 12/243,855 , Notice of Allowance, Nov. 22, 2010, 10 pages. |
| U.S. Appl. No. 12/243,855 , Non-Final Office Action, Oct. 14, 2010, 6 pages. |
| U.S. Appl. No. 12/243,855 , Notice of Allowance, Oct. 28, 2010, 5 pages. |
| U.S. Appl. No. 12/243,861 , Final Office Action, Jun. 22, 2011, 5 pages. |
| U.S. Appl. No. 12/243,861 , Non-Final Office Action, Nov. 8, 2010, 8 pgs. |
| U.S. Appl. No. 12/243,861 , Notice of Allowance, Sep. 6, 2011, 10 pgs. |
| U.S. Appl. No. 13/025,019 , Non-Final Office Action, Apr. 22, 2016, 16 pages. |
| U.S. Appl. No. 13/025,019 , Non-Final Office Action, Apr. 5, 2013, 15 pages. |
| U.S. Appl. No. 13/025,019 , Final Office Action, Aug. 28, 2015, 25 pages. |
| U.S. Appl. No. 13/025,019 , Final Office Action, Dec. 20, 2016, 16 pages. |
| U.S. Appl. No. 13/025,019 , Final Office Action, Jul. 13, 2018, 11 pages. |
| U.S. Appl. No. 15/478,048 , Final Office Action, Sep. 17, 2021, 32 pages. |
| U.S. Appl. No. 16/041,552 , Notice of Allowance, Sep. 30, 2021, 17 pages. |
| U.S. Appl. No. 16/911,154 , Non-Final Office Action, Sep. 16, 2021, 15 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20220148050 A1 | May 2022 | US |
