DETAILED ACTION
Status of the Application
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
The amendment filed on Jan 16, 2026 has been entered. The following has occurred: Claims 1, 5, 8, 15, and 19 have been amended.
Claims 1-20 are pending.
Response to Amendment
35 U.S.C. 112(b) rejection has been added.
Previous 35 U.S.C. 112(a) rejection has been withdrawn in light of the amendment, however, new 35 U.S.C. 112(a) rejection has been added.
35 U.S.C. 101 rejection has been withdrawn in light of the amendment.
35 U.S.C. 103 rejection has been maintained in light of the amendment.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for pre-AIA the inventor(s), at the time the application was filed, had possession of the claimed invention. Specifically, the examiner asserts that the Specification, as originally filled fails to disclose with enough specificity, the following limitations:
Claims 1, 8, and 15 recite “generate a digital view of the candidate object by comparing each of the received plurality of images to an orientation computer model associated with the candidate object;” which the specification lacks support. The specification in paragraph [0036], “the image is compared to an orientation model of the object. The orientation model may be a three-dimensional wireframe model of the object that is used to generate views.” In paragraph [0037], “the DA computer device may analyze the image to determine whether or not the image contains sufficient data to analyze, such as by comparing to the appropriate damage classification models.” Paragraph [0038], “the DA computer device compares the plurality of images to one or more damage classification models to determine whether the plurality of images properly display the object and damage.” Please note, the claim recites the comparison is made to an orientation computer model, not a damage classification model. The specification states the opposite of the new claim limitation. The orientation model is used to generate the view then the received images is compared to the generated view. In para. [0037]-[0038] the comparing of the image to the damage classification model is used to determine if the image contains sufficient data but not to generate a digital view. Further analysis and interpretation are provided in the 112(b) rejection below.
Claims 2-7, 9-14, and 16-20 depend from claims 1, 8, and 15 above and therefore inherit the 35 U.S.C. 112 deficiencies of their parent claim.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-20 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention.
Claims 1, 8, and 15 recite “generate a digital view of the candidate object by comparing each of the received plurality of images to an orientation computer model associated with the candidate object;” which is found to be indefinite. The function for generating a digital view by comparing received images are technologically illogical and contradicts the specification as indicated in the 112(a) rejection above. A computer cannot generate a 3D digital view by comparing 2D photos to a model. Rather, the system uses the model to generate the view and then compares the live image to the view to see if they match. For the purpose of expediting compact prosecution, the Examiner will interpret the claim limitation, “generate a digital view of the candidate object by comparing each of the received plurality of images to an orientation computer model associated with the candidate object;” to be --generate, using an orientation computer model associated with the candidate object, a digital view of the candidate object and compare each of the received plurality of images to the generated digital view--.
Claims 1 and 15 recite “continuously received” which is found to be indefinite. Per previous 112(a) rejection, the Applicant removed “continuously” from the amended claims 1 and 15 however, the other “continuously” was not removed. It becomes unclear whether if the applicant attempts to keep or remove the “continuously” from “continuously received” as part of the previous 112(a) rejection. For the purpose of expediting compact prosecution, the Examiner interprets, “continuously received” to be -- received--.
Claim 8 recites “determining selecting” which seems to be typographical error rending the claim limitation indefinite. It is unclear whether the claim limitation is supposed to be determining or selecting a repair facility. For the propose of expediting compact prosecution, the Examiner interprets “determining selecting” to be --selecting--.
Claims 1-7, 9-14, and 16-20 depend from claims 1, 8, and 15 above and therefore inherit the 35 U.S.C. 112 deficiencies of their parent claim.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under pre-AIA 35 U.S.C. 103(a) are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Taliwal et al. (US 20170293894 A1), hereinafter, “Taliwal” in view of Collins et al. (US 9824453 B1), hereinafter, “Collins,” and further in view of Mullen et al. (US 20150178852 A1), hereinafter, “Mullen.”
Claims 1, 8, and 15, Taliwal discloses a computer system, a computer-implemented method, at least one-transitory computer-readable storage medium for model-based analysis of damage to an object, the method implemented using the computer system comprising at least one processor in communication with at least one memory device, the computer-executable instructions cause the at least one processor to (Abstract: “system and method are provided for automatically estimating a repair cost for a vehicle” and para. [0006], [0045]-[0047], [0052]) for processor and memory):
train a plurality of damage classification models using historical damage data associated with damages and repairs for a plurality of objects continuously received by the computer system, wherein each of the plurality of damage classification models is configured to determine (i) an amount of damage to an object based upon a type of the object and a type of damage to the object, and (ii) how the damage would be repaired (Para. [0031], [0032], [0038], [0051], [0126], disclosing mobile phone with camera capturing images of the vehicle (i.e., object). Para. [0036], “a deep learning system (e.g., Convolutional Neural Network) is trained on a large number of images of damaged vehicles and corresponding information about damage, e.g., its extent and location on the vehicle, which are available from an insurance company's auto claims archives, in order to learn to assess damage presented with input images for a new auto claim. Such a pattern learning method can predict damage to both the exterior and interior of the vehicle, as well as the associated repair costs. The assessment of damage to the exterior determined by the image processing system can be used as input to the pattern learning system in order to supplement and refine the damage assessment. The current level of damage can be compared with the level of damage prior to filing of the current claim, as determined using image processing of prior images of the vehicle with the same system.” Para. [0037], “A comprehensive damaged parts list is then generated to prepare an estimate of the cost required to repair the vehicle by looking up in a parts database for parts and labor cost. In the absence of such a parts database, the system can be trained to predict the parts and labor cost associated with a damage assessment, since these are also available in the archival data. In some embodiments, the regions and/or areas of damage on the exterior of the vehicle can also be identified.” Para. [0121], “Some embodiments take a large number (e.g., on the order of thousands) of auto claims that contains images of the damaged vehicles and the corresponding appraisals of damaged parts, as found by auto repair shops for repair purposes. Taken together, these historical claims provide enough evidence to establish a high degree of correlation between damage visible in the images and the entire list of damaged parts, both internal and external. In one embodiment, a Convolutional Neural Network (CNN) is trained to learn this correlation. A CNN is a type of mathematical device called a neural network that can be gradually tuned to learn the patterns of correlation between its input and output from being presented a large number of exemplars of input/output pairs called training data. CNNs are configured to take into account the local structure of visual images and invariance properties of objects that are present in them. CNNs have been shown to be highly effective at the task of recognition of objects and their features provided there are enough exemplars of all possible types in the data used to train them. Some embodiments train a CNN to output a complete list of damaged parts when presented with the set of images associated to an auto claim. This includes both internal and external parts. The performance of the CNN can be made more robust when it is presented with the output of the external damage detection system described above. The output of the external damage detection system “primes” the CNN with the information about which external parts are more likely to be damaged, and thereby, increases its accuracy and speed of convergence to the solution.” Also, in para. [0138]-[0140], discloses training a machine learning system/CNNs using historical auto claims data including images and corresponding information about damages such as location on the vehicle; level of damages; interior and exterior damage prediction; which is representative evaluate damage to an object (i.e., vehicle) based upon a type of the object (e.g., vehicle make, model, color and age, damage parts) and types of damage (e.g., location of damages, interior or exterior damage, level of damages) to the object). The determined assessment for estimation of cost for part and labor is representative of determining how the damage would be repaired.);
receive, from a user computing device associated with a user (Abstract, “from a client computing device”), (i) a request for an estimate to repair a candidate object (Abstract, para. [0044], disclosing the transmitting of images for the request to estimate damage and repair cost. Also see para. [0030] and [0060]) and (ii) a plurality of images of the candidate object to repair (Abstract: “A system and method are provided for automatically estimating a repair cost for a vehicle. A method includes: receiving, at a server computing device over an electronic network, one or more images of a damaged vehicle from a client computing device; performing image processing operations on each of the one or more images to detect external damage to a first set of parts of the vehicle;” Furthermore, additional details provided in at least para. [0029]-[0032], [0051], [0060], [0126]-[0128], disclosing the use of user’s mobile phone to collect the image of the damaged vehicle to automatically calculate an estimated for repair cost);
(Claim limitation is interpreted based on the 112 rejection above) generate a digital view of the candidate object by … an orientation computer model associated with the candidate object (para. [0134], discloses displaying a 3D model of the vehicle and displaying an outline/view of the selected part for the user to capture with camera);
determine, from the properly captured images and using image recognition tools, the type of the object and the type of the damage to the candidate object (para. [0035] disclosing using a variety of computer vision techniques (i.e., image recognition tools), the system recognizes vehicle make and model (i.e., type of object) from images received and location and level of damages on exterior and interior (i.e., type of the damages) of the vehicle. Additionally, in Para. [0129], “Once the user captures the images of the damaged vehicle using the prompts provided by the vehicle claims application, the images are uploaded to a server over a network. The server is then configured to perform image processing operations on the images to identify damaged external parts, infer damaged internal parts, and estimate repair costs, as described above.” image processing operation is an example of image recognition tools to determine/identify types of damages on vehicle (i.e., object). Also see para. [0161]-[0162])),
select one or more of the plurality of trained damage classification models based upon the determined type of the object and the determined type of the damage to the candidate object (Para. [0084], “Some embodiments of the disclosure improve upon existing techniques by using a Deformable Part Model (DPM) to obtain the initial contour. DPM is a machine learning model usually used to recognize objects made of moveable parts. At a high level, DPM can be characterized by strong low-level features based on histograms of oriented gradient (HOG) that is globally invariant to illumination and locally invariant to translation and rotation, efficient matching algorithms for deformable part-based models, and discriminative learning with latent variables. After training on a large database of vehicles in various orientations, the DPM learns to put a bounding box around the vehicle in the photo. This bounding box can then serve as the initial contour.” Para. [0091], “Referring back to FIG. 5, at step 504 (i.e., image alignment), a reference image is found for the same vehicle type that is taken from the same camera position and orientation as the damaged vehicle image. Once the input image is aligned to a reference image, the server is able to overlay the two images on top of each other so that the vehicle boundaries within them more or less coincide. This is called image alignment.” Para. [0096], “At step 506 (i.e., image segmentation), the cleaned image of the damaged vehicle is segmented into vehicle parts, i.e., the boundaries of the vehicle parts are determined and drawn. Segmentation is carried out in order to assess damage on a part-by-part basis, which makes for more robust damage assessment.” Para. [0121]-[0122], [0125]-[0129] disclosing the user selects to be prompted to take images of the damaged vehicle with respect to select trained damage classification model to output damage evaluation. Specifically, in [0163] and [0189] teaching selecting the specific CNNs from a plurality of exterior part CNNs based on the predicted pose/type, wherein the image is presented only to the external part CNNs that correspond to the pose);
input, into the selected one or more trained damage classification models, the determined type of the object and the determined type of damage to the candidate object (para. [0035]-[0037], [0084], [0091], [0096], and [0129] disclosing the input of images of determined damaged vehicles and corresponding information about the damage to trained deep learning system. Specifically, para. [0189] teaching presenting the image and pose data to the selected external part CNNs);
output from the selected one or more trained damage classification models, the amount of damage to the candidate object (Claim 1 and para. [0036]-[0037], [0129], [0139],[0142]-[0144], [0162]-[0192], and [0198] disclosing the trained deep learning system (i.e. damage classification model) to assess and predict the damage of the vehicle with damaged parts list, repair cost estimation, and classifying the claim into categories such as total, medium, or small loss);
selected, based upon the amount of damage to the candidate object, a repair facility of a plurality of repair facilities to repair the damage to the candidate object (Para. [0039]: “Data aggregated across multiple claims and repair shops can also help identify misleading appraisals and recurrent fraudulent activity by repair shops. Early notification of the nature of damage can be sent to partner repair shops, allowing them to schedule the resources needed for repair early and more efficiently, reducing customer wait times, and thereby, rental vehicle costs”).
While, Taliwal suggested the sending information about the nature of damage to partner repair shops to allow schedule the resources needed for repair early and more efficiently and reduce customer wait times (Para. [0039]). However, Taliwal is not expressive on the description for determining of a time to repair the object (i.e., vehicle).
Specifically, Taliwal fails to expressly disclose (italic emphasis):
comparing each of the received plurality of images to an orientation computer model associated with the candidate object;
determine that each of the received plurality of images is properly captured by (i) matching each received image to at least a portion of the generated digital view or (ii) determining that each received image satisfies an analysis threshold associated with image acquisition parameters of each received image;
in response to determining that one of the received plurality of images fails to be properly captured, (i) generate instructions to recapture the at least one portion of the candidate object initially captured in the one of the received plurality of images and (ii) cause the user computing device to display the instructions; and in response to determining that each of the received plurality of images is properly captured:
output,… an amount of time required to repair the amount of damage to the candidate object;
transfer a data packet to a selected repair facility computer device associated with the selected repair facility, the data packet including the properly captured images; and
cause, using the data packet, the selected repair facility computer device to schedule an appointment to repair the candidate object at the selected repair facility.
However, Collins is analogous in the field of assessing vehicle damage using camera device, which specifically teaches,
comparing each of the received plurality of images to an orientation computer model associated with the candidate object (Col. 30 Ln. 44-61, “Screen 1801 may include a window 1801 a with visual aids 1801 b to aid a user in positioning an imaging device (e.g., a camera) associated with a user device (e.g., tablet, mobile phone, etc.) so that the imaging device correctly captures an image associated with instruction 1801 c. In this particular example, window 1801 a includes markers for positioning a camera so that the front and driver's side of a damaged vehicle appear in the proper location of an image.” The aligning of the live camera feed so the vehicle appears in the proper location within the markers constitutes comparing the image to the digital view);
determine that each of the received plurality of images is properly captured by (i) matching each received image to at least a portion of the generated digital view or (ii) determining that each received image satisfies an analysis threshold associated with image acquisition parameters of each received image (Col. 30 Ln. 44-61, “Screen 1801 may include a window 1801 a with visual aids 1801 b to aid a user in positioning an imaging device (e.g., a camera) associated with a user device (e.g., tablet, mobile phone, etc.) so that the imaging device correctly captures an image associated with instruction 1801 c. In this particular example, window 1801 a includes markers for positioning a camera so that the front and driver's side of a damaged vehicle appear in the proper location of an image.” Col. 36 Ln. 26-30, “Screen 2701 a may include on or more guides 2701 b. For example, screen 2701 a displays guides for orienting a user that is taking a picture from an angle that captures the front and driver-side portions of a vehicle.” teaches the matching the image to a generated digital view to ensure the proper image acquisition parameters (angles/orientation) are met);
in response to determining that one of the received plurality of images fails to be properly captured, (i) generate instructions to recapture the at least one portion of the candidate object initially captured in the one of the received plurality of images and (ii) cause the user computing device to display the instructions; and in response to determining that each of the received plurality of images is properly captured: (Col. 33 Ln. 49-65. “After server 101 receives image data transmitted from a mobile device in step 2209, server 101 (or an individual/group associated with the entity managing server 101) may determine if the photos are acceptable and the mobile device may receive, from server 101, a message indicating the acceptability of the photos in step 2211. For instance, server 101 may determine that the photos are too blurry and/or that the photos do not capture the correct angles to clearly show damage associated with the insured vehicle. If the mobile device receives a message that indicates the photos are not acceptable, the process may move to step 2213 where the server 101 may send the user instructions on what types of photos to take and/or what changes need to be made to the previously submitted photos. In this example, the process may move from step 2213 back to step 2207, where the user may take or retake photos in accordance with the received instructions.” teaches in response to determining that one of the received plurality of images fails to be properly captured (e.g., wrong angle), generating instructions to recapture the portion of the candidate object and causing the user computing device to display the instructions).
While Taliwal teaches displaying a 3D model and an outline of a vehicle part for a user to capture with a camera, Taliwal does not explicitly teach the detail of the feedback loop of rejecting the image if the user fails to properly align the camera with the outline/angle. Collins teaches using visual aids and markers to orient the user’s camera angle, and further teaches evaluating the captured photos to determine if they capture the correct angles, and if not, prompting the user with instructions to retake the photo. Therefore, it would have been obvious for one of ordinary skill in the art, before the effective filling of the invention to incorporate the automated angle-validation and recapture feedback loop of Collins with the 3D outline capture interface of model-based analysis system and method of Taliwal for the motivation of ensuring the user to comply with the 3D outline prompt and guaranteeing the images fed into the CNN/machine learning models are the correct orientation and quality, which prevents data processing errors and improves the accuracy of the automated damage assessment.
Still, the combination fails to expressly teach,
output,… an amount of time required to repair the amount of damage to the candidate object;
transfer a data packet to a selected repair facility computer device associated with the selected repair facility, the data packet including the properly captured images; and
cause, using the data packet, the selected repair facility computer device to schedule an appointment to repair the candidate object at the selected repair facility.
However, Mullen is directed to similar field of estimating the amount of damage to the vehicle and facilitating transportation of the damaged vehicle to an appropriate treatment facility. Mullen teaches,
output,… an amount of time required to repair the amount of damage to the candidate object (Claim 1 and para. [0053]-[0054] teaches the determining of complexity level that includes the estimation of repair time and cost based on the damage information of the vehicle);
(While the limitation is disclosed in Taliwal, analogous in the art, Mullen also teaches the following limitation)
selected, based upon the amount of damage to the candidate object, a repair facility of a plurality of repair facilities to repair the damage to the candidate object (Para. [0048], “one or more treatment facilities capable of performing the requisite treatment may be identified by system personnel and/or the processing center (block 208). System personnel and/or the processing center 102 may then transmit a communication related to the treatment of the damaged vehicle (block 210). For example, system personnel and/or the processing center 102 may contact one or more identified treatment facilities to initiate or inquire further in regard to the continued treatment of the damaged vehicle (block 210).” Teaching the identifying or determining of a treatment (repair) facility of the plurality of the plurality of the treatment facilities for request to repair the object/vehicle. More specifically in Fig. 6 and Para. [0055]-[0056] teaching the identifying the treatment facility out of the list of treatment facilities for treating the damaged vehicle based on prioritization of factors such as pricing structure, treatment facility capability, treatment facility location, treatment facility quality rating and/or certification, treatment facility availability, time, etc. and combinations thereof, then transmitting communication relating to the treatment to the treatment facility when identified. Similarly, in para. [0058] teaching “[a]dditional factors that may be considered when determining a repair treatment facility may include the proximity of the repair treatment facility to the damaged vehicle, e.g., collision site; the treatment facility's availability to timely repair the vehicle; and, a current or prior business relationship between the repair treatment facility and the entity using and/or administrating the treatment system 100. When the repair center is determined, information associated with the repair of the vehicle may be transmitted from system personnel and/or the processing center 102.”);
Mullen further teaches,
transfer a data packet to a selected repair facility computer device associated with the selected repair facility, the data packet including the plurality of images (Para. [0039] and [0059] teaching the transmitting of information relating to treating the damaged vehicle to a treatment facility. Fig. 1 and para. [0041]-[0042] teaching the treatment information includes images of vehicle damages from crash information and collision data are updated and stored in the system);
cause, using the data packet, the selected repair facility computer device to schedule an appointment to repair the candidate object at the selected repair facility (Para. [0048], “one or more treatment facilities capable of performing the requisite treatment may be identified by system personnel and/or the processing center (block 208). System personnel and/or the processing center 102 may then transmit a communication related to the treatment of the damaged vehicle (block 210). For example, system personnel and/or the processing center 102 may contact one or more identified treatment facilities to initiate or inquire further in regard to the continued treatment of the damaged vehicle (block 210).” Teaching the identifying or determining of a treatment (repair) facility of the plurality of the plurality of the treatment facilities for request to repair the object/vehicle. More specifically in Fig. 6 and Para. [0055]-[0056] teaching the identifying the treatment facility out of the list of treatment facilities for treating the damaged vehicle based on prioritization of factors such as pricing structure, treatment facility capability, treatment facility location, treatment facility quality rating and/or certification, treatment facility availability, time, etc. and combinations thereof, then transmitting communication relating to the treatment to the treatment facility when identified. Similarly, in para. [0058] teaching “[a]dditional factors that may be considered when determining a repair treatment facility may include the proximity of the repair treatment facility to the damaged vehicle, e.g., collision site; the treatment facility's availability to timely repair the vehicle; and, a current or prior business relationship between the repair treatment facility and the entity using and/or administrating the treatment system 100. When the repair center is determined, information associated with the repair of the vehicle may be transmitted from system personnel and/or the processing center 102. Such information may include a request to transport the damaged vehicle from the crash site directly to the repair treatment facility (block 710). The request to transport the vehicle may be sent to the selected repair treatment facility or to a vehicle transporter 114 capable of transporting damaged vehicles from collision sites.”);
Furthermore, it would have been obvious to one of ordinary skill in the art to modify in the system and method for assessing damage and repair costs in vehicles of Taliwal to include the ability to communicate with repair facility for schedule appointment as taught by Mullen, for the motivation of allowing the schedule of resources needed for repair early and more efficiently and reduce customer wait times (Taliwal para. [0085]), Further, since the claimed invention is merely a combination of old elements in a similar vehicle damage assessment for repair field of endeavor. In such combination each element merely would have performed the same vehicle damage assessment for repair related function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable (See MPEP 2143 A).
Claims 2, 9, and 16, the combination of Taliwal, Collins, and Mullen make obvious of the computer system of claim 1, the computer-implemented method of claim 8, and non-transitory computer-readable storage medium of claim 15. Taliwal further discloses,
wherein the at least one processor is further configured to determine the time to repair the candidate object based upon the amount of damage to the candidate object (Claim 1 and para. [0053]-[0054] teaches the determining of complexity level that includes the estimation of repair time and cost based on the damage information of the vehicle).
Claims 3, 10, and 17, the combination of Taliwal, Collins, and Mullen make obvious of the computer system of claim 2, the computer-implemented method of claim 9, and non-transitory computer-readable storage medium of claim 16. Mullen further teaches,
wherein the at least one processor is further configured to: compare the determined time to repair to a repair time threshold (In para. [0053] teaching “collision data may include a range of treatment complexity levels associated with various amounts of vehicle damage. In general, a treatment complexity level represents the difficulty associated with treating the damaged vehicle and may include or be associated with a pricing schema having a predetermined price structure for treating the damaged vehicle. A range of vehicle treatment complexity levels may be delineated by the amount of involvement associated with repairing and/or replacing vehicle parts of the damaged vehicle, or to scrap the damaged vehicle. Each treatment complexity level may include estimates or indications of the repair time and cost associated with the type and amount of vehicle body parts that may be damaged, e.g., body panel (front, side, rear, quarter-panel, rocker, driver-side, and passenger-side), bumper, radiator, lights, water pump, battery, struts, frame, and gas tank.” In para. [0054] teaching the comparing and matching the damaged list of vehicle parts to the collision data (which is representative of the amount of damage to the object/vehicle) to identify a treatment complexity level. Further in para. [0054] teaching “a vehicle damage estimate requiring less than 10 hours of repair time or $1000 in vehicle parts and labor may be designated as a low treatment complexity level; a vehicle damage estimate requiring between 10-15 hours of repair time or between $1000-$2500 in vehicle parts and labor may be designated as a medium treatment complexity level; a vehicle damage estimate requiring between 15-30 hours of repair time or between $2500-$5000 in vehicle parts and labor may be designated as a high treatment complexity level;” This is consistent to the description provided in the applicant’s specification in paragraph [0089], “DA computer device 410 may determine whether the time to repair exceeds a first threshold. DA computer device 410 may categorize the damage as light damage if the time to repair does not exceed the first threshold. In the exemplary embodiment, the first threshold may be 25 hours to repair. If the time to repair exceeds the first threshold, DA computer device 410 may determine whether the time to repair exceeds a second threshold. In the exemplary embodiment, the second threshold may be 49 hours to repair. If the time to repair exceeds the first threshold but not the second threshold, DA computer device 410 may categorize the damage as medium or moderate damage. If the time to repair exceeds the first threshold and the second threshold, DA computer device 410 may categorize the damage as heavy damage.” In the example provided in Mullen para. [0054], the compares repair time exceeds first threshold of 10 hours as medium treatment complexity level.); and
in response to the time to repair exceeding the repair time threshold, cause the user computing device to display an instruction indicating to take the object to the selected repair facility for obtaining an estimate for a cost to repair the candidate object (in para. [0060] of Mullen, teaching the system personnel and/or the processing center of the treatment system can transmit a request for a quote to treat the damaged vehicle from the selected treatment facilities based on the vehicle treatment complexity level. This is representative of the user computing device to instruct the user to take the object to a repair facility of the plurality of repair facilities (that is from the treatment/repair facility) for an estimated cost to repair).
The rationales to modify/combine the teachings of Taliwal with/and the teachings of Mullen are presented in the examining of independent claims 1, , and 15, and incorporated herein.
Claims 4, 11, and 18, the combination of Taliwal, Collins, and Mullen make obvious of the computer system of claim 2, the computer-implemented method of claim 9, and non-transitory computer-readable storage medium of claim 16.
Taliwal further discloses, transfer the data packet to the selected repair facility computer device, the data packet including the calculated cost to repair the candidate object (Para. [0029] disclosing an automatic vehicle damage assessment system to provide an appraisal of damage and estimate of repair cost. Para. [0037] and [0122] disclosing the estimation of repair includes cost required to repair for parts cost and labor cost. In para. [0039], “Data aggregated across multiple claims and repair shops can also help identify misleading appraisals and recurrent fraudulent activity by repair shops. Early notification of the nature of damage can be sent to partner repair shops, allowing them to schedule the resources needed for repair early and more efficiently, reducing customer wait times, and thereby, rental vehicle costs” which disclosing the information (i.e., data packet) including damages which corresponds to cost should be used to inform the repair shops).
However, Taliwal does not expressly teach,
wherein the at least one processor is further configured to: compare the determined time to repair to a repair time threshold;
in response to the time to repair not exceeding the repair time threshold, calculate a cost to repair the candidate object based upon the amount of damage to the candidate object; and
transfer the data packet to the selected repair facility computer device.
wherein the at least one processor is further configured to: compare the determined time to repair to a repair time threshold (In para. [0053] teaching “collision data may include a range of treatment complexity levels associated with various amounts of vehicle damage. In general, a treatment complexity level represents the difficulty associated with treating the damaged vehicle and may include or be associated with a pricing schema having a predetermined price structure for treating the damaged vehicle. A range of vehicle treatment complexity levels may be delineated by the amount of involvement associated with repairing and/or replacing vehicle parts of the damaged vehicle, or to scrap the damaged vehicle. Each treatment complexity level may include estimates or indications of the repair time and cost associated with the type and amount of vehicle body parts that may be damaged, e.g., body panel (front, side, rear, quarter-panel, rocker, driver-side, and passenger-side), bumper, radiator, lights, water pump, battery, struts, frame, and gas tank.” In para. [0054] teaching the comparing and matching the damaged list of vehicle parts to the collision data (which is representative of the amount of damage to the object/vehicle) to identify a treatment complexity level. Further in para. [0054] teaching “a vehicle damage estimate requiring less than 10 hours of repair time or $1000 in vehicle parts and labor may be designated as a low treatment complexity level; a vehicle damage estimate requiring between 10-15 hours of repair time or between $1000-$2500 in vehicle parts and labor may be designated as a medium treatment complexity level; a vehicle damage estimate requiring between 15-30 hours of repair time or between $2500-$5000 in vehicle parts and labor may be designated as a high treatment complexity level;” This is consistent to the description provided in the applicant’s specification in paragraph [0089], “DA computer device 410 may determine whether the time to repair exceeds a first threshold. DA computer device 410 may categorize the damage as light damage if the time to repair does not exceed the first threshold. In the exemplary embodiment, the first threshold may be 25 hours to repair. If the time to repair exceeds the first threshold, DA computer device 410 may determine whether the time to repair exceeds a second threshold. In the exemplary embodiment, the second threshold may be 49 hours to repair. If the time to repair exceeds the first threshold but not the second threshold, DA computer device 410 may categorize the damage as medium or moderate damage. If the time to repair exceeds the first threshold and the second threshold, DA computer device 410 may categorize the damage as heavy damage.” In the example provided in Mullen para. [0054], the compares repair time exceeds first threshold of 10 hours as medium treatment complexity level.);
in response to the time to repair not exceeding the repair time threshold, calculate a cost to repair the candidate object based upon the amount of damage to the candidate object (para. [0053] teaching “collision data may include a range of treatment complexity levels associated with various amounts of vehicle damage. In general, a treatment complexity level represents the difficulty associated with treating the damaged vehicle and may include or be associated with a pricing schema having a predetermined price structure for treating the damaged vehicle. A range of vehicle treatment complexity levels may be delineated by the amount of involvement associated with repairing and/or replacing vehicle parts of the damaged vehicle, or to scrap the damaged vehicle. Each treatment complexity level may include estimates or indications of the repair time and cost associated with the type and amount of vehicle body parts that may be damaged, e.g., body panel (front, side, rear, quarter-panel, rocker, driver-side, and passenger-side), bumper, radiator, lights, water pump, battery, struts, frame, and gas tank.” In para. [0054] teaching the comparing and matching the damaged list of vehicle parts to the collision data (which is representative of the amount of damage to the object/vehicle) to identify a treatment complexity level. Further in para. [0054] teaching low treatment complexity level for estimated not exceeding the threshold of 10 hours of repair time and the system provides repair cost associated with the vehicle is about $1000. This is similar to the example provided in the app. specification in paragraphs [0089], “DA computer device 410 may determine 325 a time to repair the object based upon the amount of damage. DA computer device 410 may categorize damage based upon the analysis and/or the time to repair. In some embodiments, DA computer device 410 may determine whether the time to repair exceeds a first threshold. DA computer device 410 may categorize the damage as light damage if the time to repair does not exceed the first threshold. In the exemplary embodiment, the first threshold may be 25 hours to repair. If the time to repair exceeds the first threshold, DA computer device 410 may determine whether the time to repair exceeds a second threshold.” Then in applicant’s specification paragraph [0090], “DA computer device 410 may calculate a cost to repair the object if the time to repair does not exceed the first threshold and/or the damage is categorized as light damage.” The low complexity level not exceeding 10 hours of repair time in Mullen is representative of the light damage not exceeding 25 hours to repair in the applicant’s specification); and
transfer the data packet to the selected repair facility computer device (In Fig. 6 and Para. [0055]-[0056] teaching the identifying the treatment facility for treating the damaged vehicle based on prioritization of factors such as pricing structure, treatment facility capability, treatment facility location, treatment facility quality rating and/or certification, treatment facility availability, time, etc. and combinations thereof, then transmitting communication relating to the treatment to the treatment facility when identified. Similarly, in para. [0058] teaching “[a]dditional factors that may be considered when determining a repair treatment facility may include the proximity of the repair treatment facility to the damaged vehicle, e.g., collision site; the treatment facility's availability to timely repair the vehicle; and, a current or prior business relationship between the repair treatment facility and the entity using and/or administrating the treatment system 100. When the repair center is determined, information associated with the repair of the vehicle may be transmitted from system personnel and/or the processing center 102.”)
Claims 5, 12, and 19, the combination of Taliwal, Collins, and Mullen make obvious of the computer system of claim 1, the computer-implemented method of claim 8, and non-transitory computer-readable storage medium of claim 15. Taliwal further discloses,
wherein the at least one processor is further configured to input the received plurality of images into the selected one or more trained damage classification models to provide the output (para. [0035]-[0036], [0129], [0140]-[0141] disclosing the training of a deep learning system on a large number of (input) of images of damaged vehicles and corresponding information about damage. In para. [0189], “After the pose category has been predicted by the vehicle pose classification engine 2504 for a given input image, the image is presented to each of the external part CNNs of the exterior damage detection engine 2506” to output a prediction for the damage).
Collins further teaches, wherein the image acquisition parameters include at least one of an angle, an orientation, a distance, a lighting, one or more colors, or one or more reflections of the received plurality of images (Col. 33 Ln. 49-65 teaching determining if the photos are acceptable based on the parameters such as determining if they capture the correct angles to clearly show damage associated with the insured vehicle and use visual guides to ensure proper camera positioning and orientation).
The rationales to modify/combine the teachings of Taliwal with/and the teachings of Collins are presented in the examining of independent claims 1, 8, and 15, and incorporated herein.
Claims 6 and 13, the combination of Taliwal, Collins, and Mullen make obvious of the computer system of claim 1 and the computer-implemented method of claim 8. Taliwal further discloses,
wherein the historical damage data includes at least one of historical images, historical estimates, or historical repair costs (Para. [0121], “Some embodiments take a large number (e.g., on the order of thousands) of auto claims that contains images of the damaged vehicles and the corresponding appraisals of damaged parts, as found by auto repair shops for repair purposes. Taken together, these historical claims provide enough evidence to establish a high degree of correlation between damage visible in the images and the entire list of damaged parts, both internal and external”, disclosing historical images. Para. [0139], “Some embodiments use machine learning to perform the task of prediction of vehicle damage from an auto claim. Thousands of historical auto claims are stored in one or more databases, such as database 110 in FIG. 1, for training and testing of the disclosed system. The database also stored auto claim images and other pieces of information that come with a claim, such as vehicle make, model, color, age, and current market value, for example. The desired output of the disclosed machine learning system is the damage appraisal as prepared by a repair shop consisting of a list of parts that were repaired or replaced and the corresponding costs of repair, both for parts and labor.” Which disclosing historical images from auto claim and historical estimates of value and cost of repair).
Claims 7, 14, and 20, the combination of Taliwal, Collins, and Mullen make obvious of the computer system of claim 1, the computer-implemented method of claim 8, and non-transitory computer-readable storage medium of claim 15. Taliwal further discloses,
wherein the plurality of damage classification models are configured to simulate the damage to the object and repairs necessary to fix the damage to the object (Para. [0036], “a deep learning system (e.g., Convolutional Neural Network) is trained on a large number of images of damaged vehicles and corresponding information about damage, e.g., its extent and location on the vehicle, which are available from an insurance company's auto claims archives, in order to learn to assess damage presented with input images for a new auto claim. Such a pattern learning method can predict damage to both the exterior and interior of the vehicle, as well as the associated repair costs. The assessment of damage to the exterior determined by the image processing system can be used as input to the pattern learning system in order to supplement and refine the damage assessment. The current level of damage can be compared with the level of damage prior to filing of the current claim, as determined using image processing of prior images of the vehicle with the same system.” Para. [0037], “A comprehensive damaged parts list is then generated to prepare an estimate of the cost required to repair the vehicle by looking up in a parts database for parts and labor cost. In the absence of such a parts database, the system can be trained to predict the parts and labor cost associated with a damage assessment, since these are also available in the archival data. In some embodiments, the regions and/or areas of damage on the exterior of the vehicle can also be identified.” Para. [0121], “Some embodiments take a large number (e.g., on the order of thousands) of auto claims that contains images of the damaged vehicles and the corresponding appraisals of damaged parts, as found by auto repair shops for repair purposes. Taken together, these historical claims provide enough evidence to establish a high degree of correlation between damage visible in the images and the entire list of damaged parts, both internal and external. In one embodiment, a Convolutional Neural Network (CNN) is trained to learn this correlation. A CNN is a type of mathematical device called a neural network that can be gradually tuned to learn the patterns of correlation between its input and output from being presented a large number of exemplars of input/output pairs called training data. CNNs are configured to take into account the local structure of visual images and invariance properties of objects that are present in them. CNNs have been shown to be highly effective at the task of recognition of objects and their features provided there are enough exemplars of all possible types in the data used to train them. Some embodiments train a CNN to output a complete list of damaged parts when presented with the set of images associated to an auto claim. This includes both internal and external parts. The performance of the CNN can be made more robust when it is presented with the output of the external damage detection system described above. The output of the external damage detection system “primes” the CNN with the information about which external parts are more likely to be damaged, and thereby, increases its accuracy and speed of convergence to the solution.” Para. [0138], “In another implementation of the automatic vehicle damage assessment (AVDA) system, rather than comparing photos of a damaged vehicle to an undamaged version, another embodiment of the disclosure relies upon machine learning methods to learn patterns of vehicle damage from a large number of auto claims in order to predict damage for a new claim. In general, machine learning systems are systems that use “training data” to “learn” to associate their input with a desired output. Learning is done by changing parameters of the system until the system outputs results as close to the desired outputs as possible. Once such a machine system has learned the input-output relationship from the training data, the machine learning system can be used to predict the output upon receiving a new input for which the output may not be known. The larger the training data set and the more representative of the input space, the better the machine learning system performs on the prediction task.” Para. [0139], “Some embodiments use machine learning to perform the task of prediction of vehicle damage from an auto claim. Thousands of historical auto claims are stored in one or more databases, such as database 110 in FIG. 1, for training and testing of the disclosed system. The database also stored auto claim images and other pieces of information that come with a claim, such as vehicle make, model, color, age, and current market value, for example. The desired output of the disclosed machine learning system is the damage appraisal as prepared by a repair shop consisting of a list of parts that were repaired or replaced and the corresponding costs of repair, both for parts and labor. Another desired output is the determination of the loss type, namely, total loss, medium loss, or small loss, for example.”)
Response to Remarks
35 U.S.C. 101 Rejections:
Notwithstanding the Applicant’s remarks, the amended claim integrates the abstract idea into a practical application. The claimed invention no longer recites just appraising damage on a computer. Instead, the amended claim recites a specific technical image-validation feedback loop of using a 3D orientation computer model to generate digital views, matches user submitted/inputted images with the generated digital views and automatically determine/validate the submitted images to be recaptured again if they fail to be properly captured (poorly framed) before feeding the data into the machine learning mode. The pre-processing validation step solves or improves the technical problem of automated image analysis with the application of machine learning model. Therefore, the 101 rejection is withdrawn.
35 U.S.C. 103 Rejections:
The Examiner asserts that the Applicant’s arguments are directed towards amended claim limitations and are, therefore, considered moot. However, the Examiner has responded to the amended amendments, which the arguments are directed to, in the rejection above, thereby addressing the Applicant’s arguments. New reference, Collins have been introduced to teach the newly amended claim limitations.
Relevant Prior Art Not Relied Upon
The prior art made of record and not relied upon is considered pertinent to Applicant’s disclosure. The additional cited art, including but not limited to the excerpts below, further establishes the state of the art at the time of Applicant’s invention and shows the following was known:
Liang Wei and Yu Haibin, ("Learning based dynamic approach to job-shop scheduling," 2001 International Conferences on Info-Tech and Info-Net. Proceedings (Cat. No.01EX479), Beijing, China, 2001, pp. 274-279 vol.3, doi: 10.1109/ICII.2001.983069.) which teaches the use of machine learning for scheduling candidate to simulate the problem.
Zhang et al. (US 20180293552 A1) is directed to methods, systems, and computer-readable storage media for generation of a vehicle repair plan. Implementations include actions of receiving vehicle damage data including an image of a damaged vehicle. The vehicle damage data is processed to determine a first vehicle component. The first image is matched to a second image to determine a second vehicle component within the second image. The second vehicle component is processed to determine a damaged area and a damage type of a portion of the damaged vehicle. A maintenance plan is generated for the damaged vehicle based on the damaged area and the damage type. The maintenance plan is initiated for the damaged vehicle.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to WENREN CHEN whose telephone number is (571)272-5208. The examiner can normally be reached Monday - Friday 10AM - 6PM.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Nathan C Uber can be reached on (571) 270-3923. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/WENREN CHEN/Examiner, Art Unit 3626
Prosecution Insights