Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
2. This communication is responsive to Application No. 18/973,589 and the amendments filed on 6/22/2026.
3. Claims 1, 3-4, and 6-18 are presented for examination.
Response to Arguments
4. Applicant’s arguments, see pages 11-12, filed 6/22/2026, with respect to the objection to claim 3 for minor informalities have been fully considered and are persuasive. The objection of 3/20/2026 has been withdrawn.
5. Applicant’s arguments, see page 12, filed 6/22/2026, with respect to the rejection of claim 20 under 35 U.S.C. 112(b) have been fully considered and are persuasive. The rejection of claim 20 under 35 U.S.C. 112(b) of 3/20/2026 has been withdrawn.
6. Applicant’s arguments with respect to the rejection of claim(s) 1-20 under 35 U.S.C. 103 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Regarding independent claim 1, the Examiner agrees that the combination of US 20220297297 A1 to Huang and “Dimensionality reduction for hand-independent dexterous robotic grasping” to Ciocarlie fails to teach all of the limitations of amended claim 1. However, in light of the amendments and the Applicant’s remarks, an updated search was conducted, and a new ground of rejection concerning claim 1 has been determined, in which will be described later.
Regarding claims 2, 5, 19, and 20, these claims have been cancelled, and thus, are withdrawn from further consideration.
Regarding dependent claims 3-4 and 6-17, as all of these claims depend from claim 1, are still rejected, in which will be described later.
Regarding independent claim 18, as this claim contains similar limitations as claim 1, is still rejected for similar reasons as claim 1 is, in which will be described later.
7. Examiner notes Applicant's request for interview, if any issues remain that would prevent allowance of the application, however, in light of the new ground of rejection in view of the newly found prior art, which was necessitated based on Applicant's amendments to the claims, Examiner would like to provide the Applicant with an opportunity to review the new grounds of rejection in view of said newly found prior art, prior to an interview. Once the Applicant has reviewed the new grounds of rejection, Examiner is willing to conduct an interview to discuss any remaining issues with the application that Applicant would like to discuss.
Claim Rejections - 35 USC § 103
8. 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 (i.e., changing from AIA to pre-AIA ) 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.
9. 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.
10. Claim(s) 1, 3, 6, and 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602).
Regarding Claim 1, Huang teaches a method of grasp generation for a robot, the method comprising: searching within the subspace of the configuration space of the robot hand model for a plurality of robot hand configurations to engage the target graspable part of the object model with the grasp type ([0060] via “A physics simulator 110 receives a robotic grasper model 108 and deformable object models 104 from a storage system 102. The physics simulator 110 also receives grasping models 116 to apply to the robotic grasper model 108
and deformable object models 104. The physics simulator 110 simulates effects of the various grasping models 116 on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116.”), (Note: The Examiner interprets the robotic grasper model 108 of Huang as the robot hand model.), the searching comprising minimizing contact energy between the robot hand model and the object model, wherein the contact energy depends at least in part on a distance between the set of robot contacts and the target graspable part ([0053] via “In one embodiment the robotic grasp features comprise seven metrics that partially encode the performance metrics. These features may be determined/applied before the robotic grasper picks up the object, but after applying the desired grasping force with the robotic grasper. … In one embodiment, the utilized robotic grasp features are pure distance, perpendicular distance, number of contacts, edge distance, squeeze distance, robotic grasper separation, and gravity alignment.”), ([0054] via “Pure distance refers to the distance from the center of each robotic grasper finger's contact patch to the object's center of mass (COM), averaged over the number of fingers. Perpendicular distance is the average of the perpendicular distance from the center of each finger's contact patch to the COM.”), ([0069] via “In some implementations, for prism-shaped objects the pure distance feature may be assigned the highest importance or dominant status among the grasp performance metrics. … Perpendicular distance is an additional important feature for angular instability on prism-shaped objects, and is the most important robotic grasp feature for spheroid-shaped objects.”);
generating a set of candidate grasps based on the plurality of robot hand configurations ([0060] via “The physics simulator 110 also receives grasping models 116 to apply to the robotic grasper model 108 and deformable object models 104.”), ([0077] via “FIG. 3 depicts a process 300 for configuring control settings in a robotic grasper in one embodiment. For each grasps model of a plurality of candidate grasping models of a deformable object, a simulation (block 302) is generated of a response of the deformable object to the grasping model.”);
simulating grasping of the object model with the robot hand model a plurality of times for a given candidate grasp from the set of candidate grasps ([0060] via “The physics simulator
110 also receives grasping models 116 to apply to the robotic grasper model 108 and deformable object models 104. The physics simulator 110 simulates effects of the various grasping models 116 on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116.”), wherein each simulating grasping of the object model comprises:
generating a simulated grasp based on the given candidate grasp ([0060] via “The physics simulator 110 also receives grasping models 116 to apply to the robotic grasper model
108 and deformable object models 104. The physics simulator 110 simulates effects of the various grasping models 116 on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116.”);
executing the simulated grasp in a physics engine to cause the robot hand model to engage the object model with the simulated grasp ([0060] via “The physics simulator 110 also receives grasping models 116 to apply to the robotic grasper model 108 and deformable object models 104. The physics simulator 110 simulates effects of the various grasping models 116 on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116.”), ([0075] via “In block 202, a physics simulator in operated on a deformable object model to generate a simulation of a response of the deformable object model to a grasping model for the robotic manipulator.”), (Note: See Figures 5 and 7-14 of Huang as well.);
applying a wrench disturbance to the object model while the robot hand model engages the object model with the simulated grasp in the physics engine ([0044] via “Grasp performance metrics are generated utilizing a (e.g., tetrahedral) mesh representation of an object (an object model) and a set of candidate grasp object models for simulation. Robotic grasp outcomes from the simulations are quantified by metrics such as stress, deformation, and stability.”), ([0075] via “In block 202, a physics simulator in operated on a deformable object model to generate a simulation of a response of the deformable object model to a grasping model for the robotic manipulator. In block 204, grasp performance metrics are obtained from the simulation for stress, deformation, controllability, and instability of the response to the grasping model.”), (Note: See Figures 7-14 of Huang as well. The Examiner interprets at least the stress (i.e., forces) of the object model of Huang as the wrench disturbance, as the wrench disturbance is defined in paragraphs [0140] and [0167] of the specification of the instant application.);
measuring a response of the object model to the applied wrench disturbance ([0044] via “Robotic grasp outcomes from the simulations are quantified by metrics such as stress, deformation, and stability. … In one embodiment, GPU-based Finite Element Method simulation is utilized to efficiently simulate a robust set of robotic grasp object models on diverse object categories and to quantify (via metrics) a quality of the grasp object models.”), ([0075] via “In block 204, grasp performance metrics are obtained from the simulation for stress, deformation, controllability, and instability of the response to the grasping model. In block 206, the grasp performance metrics are transformed into robotic grasp features. In block 208, it is determined whether a conforming model (of the grasping models) comprises robotic grasp features that meet manipulation constraints of the deformable object.”), (Note: See Figures 7-14 of Huang as well.); and
assigning a grasp stability score to the simulated grasp based on the measured response ([0044] via “Grasp performance metrics are generated utilizing a (e.g., tetrahedral) mesh representation of an object (an object model) and a set of candidate grasp object models for simulation. Robotic grasp outcomes from the simulations are quantified by metrics such as stress, deformation, and stability. … In one embodiment, GPU-based Finite Element Method simulation is utilized to efficiently simulate a robust set of robotic grasp object models on diverse object categories and to quantify (via metrics) a quality of the grasp object models.”), ([0075] via “In block 208, it is determined whether a conforming model (of the grasping models) comprises robotic grasp features that meet manipulation constraints of the deformable object.”); and
generating a set of feasible grasps for the given candidate grasp based on the respective simulated grasps having individual grasp stability scores above a grasp stability score threshold at a target wrench disturbance ([0062] via “Robotic grasp features 120 correspond to the robotic grasp features 118 for one of the grasping models 116 best meeting the manipulation constraints 122 may selected and applied to the path planner 114 for a robotic grasper 124 to be utilized on the physical object in the application.”), ([0075] via “In block 204, grasp performance metrics are obtained from the simulation for stress, deformation, controllability, and instability of the response to the grasping model. In block 206, the grasp performance metrics are transformed into robotic grasp features. … In block 210, if (on condition that) the robotic grasp features from the simulation meet the manipulation constraints, a path planner for the robotic manipulator is configured with grasp settings associated with the grasping model.”).
Huang is silent on determining a subspace of a configuration space of a robot hand model having a set of degrees of freedom based on a set of eigengrasps for a grasp type, wherein each eigengrasp defines a subset of the set of degrees of freedom that is movable for the grasp type; attaching a target graspable part to an object model to define an allowable contact area on the object model based on a grasping context; and attaching a set of robot contacts to the robot hand model to define allowable contact points on the robot hand model based on the grasp type.
However, Ciocarlie teaches determining a subspace of a configuration space of a robot hand model having a set of degrees of freedom (Section II paragraph 1 via “If d is the number of degrees of freedom (DOF) of the hand, than a posture p can be defined as … where θi is the value of i-th degree of freedom.”), (Note: See Table 1 and Figure 3 of Ciocarlie as well.) based on a set of eigengrasps for a grasp type (Section II paragraphs 3-4 via “In our work, we will refer to the Principal Components of these postures as eigengrasps. The implication is that they form a low-dimensionality basis for grasp postures, and can be linearly combined to closely approximate most common grasping positions. Each eigengrasp ei is a d-dimensional vector and can also be thought of as direction of motion in joint space. Motion along one eigengrasp direction will usually imply motion along all (or most) degrees of freedom of the hand. … By choosing a basis comprising b eigengrasps, a hand posture placed in the subspace defined by this basis can be expressed as a function of the amplitudes ai along each eigengrasp direction: … and is therefore completely defined by the amplitudes vector a = [a1 . . . ab] ∈ Rb.”), wherein each eigengrasp defines a subset of the set of degrees of freedom that is movable for the grasp type (Section III A paragraph 2 via “We therefore propose performing the optimization in eigengrasp space, as opposed to DOF space. The energy function takes the form … where a ∈ R2 is the vector of eigengrasp amplitudes. This effectively reduces the parameter space to 8 dimensions (2 eigengrasp amplitudes plus 6 extrinsic DOF’s) ….”), (Section V paragraph 1 via “For each model, we have defined a low dimensional subspace of the degrees of freedom space, determined by a number of basis vectors which we call eigengrasps.”).
Further, Fan ‘602 teaches attaching a target graspable part to an object model to define an allowable contact area on the object model based on a grasping context ([0045] via “Sampling points are also defined on the surface of the object 330 (including the target grasp region). FIG. 3 is illustrated as a 2D cross-section for clarity.”), ([0046] via “For each of the points p.sub.i (324) on the gripper parts 320/322, a nearest neighbor point 334 on the surface of the object 330 is determined. … From the points p.sub.i on the gripper parts 320/322 and the corresponding nearest neighbor points, an optimization computation is performed to update the gripper translation, rotation and opening so that the grasp quality is improved and interference is reduced.”), (Note: See Figures 3 and 4 of Fan ‘602 as well. The Examiner interprets the points on the object of Fan ‘602 as the target graspable part.); and
attaching a set of robot contacts to the robot hand model to define allowable contact points on the robot hand model based on the grasp type ([0045] via “A set of points p.sub.i is defined on the surfaces of the gripper parts 320/322, using any suitable sampling density. The points p.sub.i are defined on any portion of the gripper parts 320/322 (i.e., the tips of the fingers 222 of the gripper model 220) which may contact the object 330 for grasping; this is defined at the step 230 of FIG. 2 by user selection of gripper contact surfaces (such as triangular patches on the surface of the fingers 222 of the gripper model 220).”), ([0046] via “For each of the points p.sub.i (324) on the gripper parts 320/322, a nearest neighbor point
334 on the surface of the object 330 is determined. ... From the points p.sub.i on the gripper parts 320/322 and the corresponding nearest neighbor points, an optimization computation is performed to update the gripper translation, rotation and opening so that the grasp quality is improved and interference is reduced.”), (Note: See Figures 3 and 4 of Fan ‘602 as well. The Examiner interprets the points on the gripper of Fan ‘602 as the robot contacts.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Ciocarlie wherein the method comprises: determining a subspace of a configuration space of a robot hand model having a set of degrees of freedom based on a set of eigengrasps for a grasp type, wherein each eigengrasp defines a subset of the set of degrees of freedom that is movable for the grasp type. Doing so results in faster processing of the simulations and reduces the complexity of the simulated space, as stated by Ciocarlie (Section III B paragraph 3 via “The eigengrasp concept allows us to design flexible control algorithms that operate identically across all the presented hand models. The key to our approach is that the eigengrasps encapsulate the kinematic characteristics of each hand design. This enables control algorithms that operate on eigengrasp amplitudes to ignore low-level operations and concentrate on the high-level task. … Another advantage is the significant dimensionality reduction (by as much as a factor of 10 for complex hands) obtained by operating in the reduced basis eigengrasp space as opposed to the full joint space.”).
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Fan ‘602 wherein the method comprises: attaching a target graspable part to an object model to define an allowable contact area on the object model based on a grasping context; and attaching a set of robot contacts to the robot hand model to define allowable contact points on the robot hand model based on the grasp type. Doing so quickly and effectively identifies feasible grasps for the robot hand model with respect to the object model by shrinking the models to their respective graspable regions, as stated by Fan ‘602 ([0076] via “The region-based grasp generation technique shown in FIGS. 2-4 has been demonstrated on a variety of real-world parts to produce a high quality grasp database—including a diversity of grasp approach directions and grasp locations on the part, while using only the target grasp regions defined by the user. The computed grasp database has also been demonstrated to enable quick and effective grasp identification in a robotic grasping system of the type shown in FIG. 1.”).
Regarding Claim 3, modified reference Huang teaches the method of claim 1, wherein searching within the subspace of the configuration space of the robot hand model for the plurality of robot hand configurations further comprises minimizing a weighted sum of forces and/or torques produced at contact points between the robot hand model and the object model ([0047] via “In one embodiment a set of grasp performance metrics is generated comprising grasp success, stress, deformation, strain energy, linear and angular instability, and deformation controllability.”), ([0048] via “Stress is the element-wise stress distribution of an object when picked up. Exceeding material stress thresholds (e.g., yield stress, ultimate stress) leads to permanent deformation, damage, or fracture. … Each element's stress tensor may be transformed into von Mises stress, a scalar quantity that quantifies whether an element has exceeded its yield threshold. … One embodiment measures the maximum von Mises stress due to real-world constraints for a plurality of scenarios to avoid damage to objects being manipulated.”).
Regarding Claim 6, modified reference Huang teaches the method of claim 1, wherein generating the set of candidate grasps based on the plurality of robot hand configurations comprises: determining a grasp quality score for each of the robot hand configurations based on a set of grasp quality metrics ([0075] via “In block 208, it is determined whether a conforming model (of the grasping models) comprises robotic grasp features that meet manipulation constraints of the deformable object.”); and
generating the set of candidate grasps based on the robot hand configurations having quality scores above a grasp quality score threshold ([0062] via “Robotic grasp features 120
correspond to the robotic grasp features 118 for one of the grasping models 116 best meeting the manipulation constraints 122 may selected and applied to the path planner 114 for a robotic grasper 124 to be utilized on the physical object in the application.”), ([0075] via “In block 204, grasp performance metrics are obtained from the simulation for stress, deformation, controllability, and instability of the response to the grasping model. In block
206, the grasp performance metrics are transformed into robotic grasp features. … In block
210, if (on condition that) the robotic grasp features from the simulation meet the manipulation constraints, a path planner for the robotic manipulator is configured with grasp settings associated with the grasping model.”).
Regarding Claim 18, Huang teaches one or more non-transitory computer-readable storage media storing computer-executable instructions ([0079] via “In one embodiment, a computing apparatus to carry out the process 300 and/or process 400 includes at least one processor (e.g., one or more general purpose data processor and one or more graphics processing unit). The computing apparatus also includes a memory storing instructions that, when executed by the at least one processor, configure the apparatus to ….”), ([0275] via “Logic” refers to machine memory circuits and non-transitory machine readable media comprising machine-executable instructions (software and firmware), and/or circuitry (hardware) … that may be applied to influence the operation of a device.”) that when executed perform operations comprising:
searching within the subspace of the configuration space of the robot hand model for a plurality of robot hand configurations to engage the target graspable part of the object model with the grasp type ([0060] via “A physics simulator 110 receives a robotic grasper model 108
and deformable object models 104 from a storage system 102. The physics simulator 110 also receives grasping models 116 to apply to the robotic grasper model 108 and deformable object models 104. The physics simulator 110 simulates effects of the various grasping models 116 on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116.”), (Note: The Examiner interprets the robotic grasper model 108 of Huang as the robot hand model.”), the searching comprising minimizing contact energy between the robot hand model and the object model, wherein the contact energy depends at least in part on a distance between the set of robot contacts and the target graspable part ([0053] via “In one embodiment the robotic grasp features comprise seven metrics that partially encode the performance metrics. These features may be determined/applied before the robotic grasper picks up the object, but after applying the desired grasping force with the robotic grasper. … In one embodiment, the utilized robotic grasp features are pure distance, perpendicular distance, number of contacts, edge distance, squeeze distance, robotic grasper separation, and gravity alignment.”), ([0054] via “Pure distance refers to the distance from the center of each robotic grasper finger's contact patch to the object's center of mass (COM), averaged over the number of fingers. Perpendicular distance is the average of the perpendicular distance from the center of each finger's contact patch to the COM.”), ([0069] via “In some implementations, for prism-shaped objects the pure distance feature may be assigned the highest importance or dominant status among the grasp performance metrics. … Perpendicular distance is an additional important feature for angular instability on prism-shaped objects, and is the most important robotic grasp feature for spheroid-shaped objects.”);
generating a set of candidate grasps based on the plurality of robot hand configurations ([0060] via “The physics simulator 110 also receives grasping models 116 to apply to the robotic grasper model 108 and deformable object models 104.”), ([0077] via “FIG. 3 depicts a process 300 for configuring control settings in a robotic grasper in one embodiment. For each grasps model of a plurality of candidate grasping models of a deformable object, a simulation (block 302) is generated of a response of the deformable object to the grasping model.”);
simulating grasping of the object model with the robot hand model a plurality of times for a given candidate grasp from the set of candidate grasps ([0060] via “The physics simulator
110 also receives grasping models 116 to apply to the robotic grasper model 108 and deformable object models 104. The physics simulator 110 simulates effects of the various grasping models 116 on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116.”), wherein each simulating grasping of the object model comprises:
generating a simulated grasp based on the given candidate grasp ([0060] via “The physics simulator 110 also receives grasping models 116 to apply to the robotic grasper model 108 and deformable object models 104. The physics simulator 110 simulates effects of the various grasping models 116 on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116.”);
executing the simulated grasp in a physics engine to cause the robot hand model to engage the object model with the simulated grasp ([0060] via “The physics simulator 110 also receives grasping models 116 to apply to the robotic grasper model 108 and deformable object models 104. The physics simulator 110 simulates effects of the various grasping models 116 on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116.”), ([0075] via “In block 202, a physics simulator in operated on a deformable object model to generate a simulation of a response of the deformable object model to a grasping model for the robotic manipulator.”), (Note: See Figures 5 and 7-14 of Huang as well.);
applying a wrench disturbance to the object model while the robot hand model engages the object model with the simulated grasp in the physics engine ([0044] via “Grasp performance metrics are generated utilizing a (e.g., tetrahedral) mesh representation of an object (an object model) and a set of candidate grasp object models for simulation. Robotic grasp outcomes from the simulations are quantified by metrics such as stress, deformation, and stability.”), ([0075] via “In block 202, a physics simulator in operated on a deformable object model to generate a simulation of a response of the deformable object model to a grasping model for the robotic manipulator. In block 204, grasp performance metrics are obtained from the simulation for stress, deformation, controllability, and instability of the response to the grasping model.”), (Note: See Figures 7-14 of Huang as well. The Examiner interprets at least the stress (i.e., forces) of the object model of Huang as the wrench disturbance, as the wrench disturbance is defined in paragraphs [0140] and [0167] of the specification of the instant application.);
measuring a response of the object model to the applied wrench disturbance ([0044] via “Robotic grasp outcomes from the simulations are quantified by metrics such as stress, deformation, and stability. … In one embodiment, GPU-based Finite Element Method simulation is utilized to efficiently simulate a robust set of robotic grasp object models on diverse object categories and to quantify (via metrics) a quality of the grasp object models.”), ([0075] via “In block 204, grasp performance metrics are obtained from the simulation for stress, deformation, controllability, and instability of the response to the grasping model. In block 206, the grasp performance metrics are transformed into robotic grasp features. In block 208, it is determined whether a conforming model (of the grasping models) comprises robotic grasp features that meet manipulation constraints of the deformable object.”), (Note: See Figures 7-14 of Huang as well.); and
assigning a grasp stability score to the simulated grasp based on the measured response ([0044] via “Grasp performance metrics are generated utilizing a (e.g., tetrahedral) mesh representation of an object (an object model) and a set of candidate grasp object models for simulation. Robotic grasp outcomes from the simulations are quantified by metrics such as stress, deformation, and stability. … In one embodiment, GPU-based Finite Element Method simulation is utilized to efficiently simulate a robust set of robotic grasp object models on diverse object categories and to quantify (via metrics) a quality of the grasp object models.”), ([0075] via “In block 208, it is determined whether a conforming model (of the grasping models) comprises robotic grasp features that meet manipulation constraints of the deformable object.”); and
generating a set of feasible grasps for the given candidate grasp based on the respective simulated grasps having individual grasp stability scores above a grasp stability score threshold at a target wrench disturbance ([0062] via “Robotic grasp features 120 correspond to the robotic grasp features 118 for one of the grasping models 116 best meeting the manipulation constraints 122 may selected and applied to the path planner 114 for a robotic grasper 124 to be utilized on the physical object in the application.”), ([0075] via “In block 204, grasp performance metrics are obtained from the simulation for stress, deformation, controllability, and instability of the response to the grasping model. In block 206, the grasp performance metrics are transformed into robotic grasp features. … In block 210, if (on condition that) the robotic grasp features from the simulation meet the manipulation constraints, a path planner for the robotic manipulator is configured with grasp settings associated with the grasping model.”).
Huang is silent on determining a subspace of a configuration space of a robot hand model having a set of degrees of freedom based on a set of eigengrasps for a grasp type, wherein each eigengrasp defines a subset of the set of degrees of freedom that is movable for the grasp type; attaching a target graspable part to an object model to define an allowable contact area of the object model based on a grasping context; and attaching a set of robot contacts to the robot hand model to define allowable contact points on the robot hand model based on the grasp type.
However, Ciocarlie teaches determining a subspace of a configuration space of a robot hand model having a set of degrees of freedom (Section II paragraph 1 via “If d is the number of degrees of freedom (DOF) of the hand, than a posture p can be defined as … where θi is the value of i-th degree of freedom.”), (Note: See Table 1 and Figure 3 of Ciocarlie as well.) based on a set of eigengrasps for a grasp type (Section II paragraphs 3-4 via “In our work, we will refer to the Principal Components of these postures as eigengrasps. The implication is that they form a low-dimensionality basis for grasp postures, and can be linearly combined to closely approximate most common grasping positions. Each eigengrasp ei is a d-dimensional vector and can also be thought of as direction of motion in joint space. Motion along one eigengrasp direction will usually imply motion along all (or most) degrees of freedom of the hand. … By choosing a basis comprising b eigengrasps, a hand posture placed in the subspace defined by this basis can be expressed as a function of the amplitudes ai along each eigengrasp direction: … and is therefore completely defined by the amplitudes vector a = [a1 . . . ab] ∈ Rb.”), wherein each eigengrasp defines a subset of the set of degrees of freedom that is movable for the grasp type (Section III A paragraph 2 via “We therefore propose performing the optimization in eigengrasp space, as opposed to DOF space. The energy function takes the form … where a ∈ R2 is the vector of eigengrasp amplitudes. This effectively reduces the parameter space to 8 dimensions (2 eigengrasp amplitudes plus 6 extrinsic DOF’s) ….”), (Section V paragraph 1 via “For each model, we have defined a low dimensional subspace of the degrees of freedom space, determined by a number of basis vectors which we call eigengrasps.”).
Further, Fan ‘602 teaches attaching a target graspable part to an object model to define an allowable contact area of the object model based on a grasping context ([0045] via “Sampling points are also defined on the surface of the object 330 (including the target grasp region). FIG. 3 is illustrated as a 2D cross-section for clarity.”), ([0046] via “For each of the points p.sub.i (324) on the gripper parts 320/322, a nearest neighbor point 334 on the surface of the object 330 is determined. … From the points p.sub.i on the gripper parts 320/322 and the corresponding nearest neighbor points, an optimization computation is performed to update the gripper translation, rotation and opening so that the grasp quality is improved and interference is reduced.”), (Note: See Figures 3 and 4 of Fan ‘602 as well. The Examiner interprets the points on the object of Fan ‘602 as the target graspable part.); and
attaching a set of robot contacts to the robot hand model to define allowable contact points on the robot hand model based on the grasp type ([0045] via “A set of points p.sub.i is defined on the surfaces of the gripper parts 320/322, using any suitable sampling density. The points p.sub.i are defined on any portion of the gripper parts 320/322 (i.e., the tips of the fingers 222 of the gripper model 220) which may contact the object 330 for grasping; this is defined at the step 230 of FIG. 2 by user selection of gripper contact surfaces (such as triangular patches on the surface of the fingers 222 of the gripper model 220).”), ([0046] via “For each of the points p.sub.i (324) on the gripper parts 320/322, a nearest neighbor point
334 on the surface of the object 330 is determined. ... From the points p.sub.i on the gripper parts 320/322 and the corresponding nearest neighbor points, an optimization computation is performed to update the gripper translation, rotation and opening so that the grasp quality is improved and interference is reduced.”), (Note: See Figures 3 and 4 of Fan ‘602 as well. The Examiner interprets the points on the gripper of Fan ‘602 as the robot contacts.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Ciocarlie wherein the operations comprise: determining a subspace of a configuration space of a robot hand model having a set of degrees of freedom based on a set of eigengrasps for a grasp type, wherein each eigengrasp defines a subset of the set of degrees of freedom that is movable for the grasp type. Doing so results in faster processing of the simulations and reduces the complexity of the simulated space, as stated by Ciocarlie (Section III B paragraph 3 via “The eigengrasp concept allows us to design flexible control algorithms that operate identically across all the presented hand models. The key to our approach is that the eigengrasps encapsulate the kinematic characteristics of each hand design. This enables control algorithms that operate on eigengrasp amplitudes to ignore low-level operations and concentrate on the high-level task. … Another advantage is the significant dimensionality reduction (by as much as a factor of 10 for complex hands) obtained by operating in the reduced basis eigengrasp space as opposed to the full joint space.”).
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Fan ‘602 wherein the operations comprise: attaching a target graspable part to an object model to define an allowable contact area of the object model based on a grasping context; and attaching a set of robot contacts to the robot hand model to define allowable contact points on the robot hand model based on the grasp type. Doing so quickly and effectively identifies feasible grasps for the robot hand model with respect to the object model by shrinking the models to their respective graspable regions, as stated by Fan ‘602 ([0076] via “The region-based grasp generation technique shown in FIGS. 2-4 has been demonstrated on a variety of real-world parts to produce a high quality grasp database—including a diversity of grasp approach directions and grasp locations on the part, while using only the target grasp regions defined by the user. The computed grasp database has also been demonstrated to enable quick and effective grasp identification in a robotic grasping system of the type shown in FIG. 1.”).
11. Claim(s) 4 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), and further in view of Fan '599 (US 20230124599 A1 hereinafter Fan '599) and Petersen et al. (US 20250196333 A1 hereinafter Petersen).
Regarding Claim 4, modified reference Huang teaches the method of claim 1, but is silent on wherein searching within the subspace of the configuration space of the robot hand model for the plurality of robot hand configurations comprises minimizing penetration of the robot hand model into the object model and minimizing self-collision of the robot hand model.
However, Fan ‘599 teaches minimizing penetration of the robot hand model into the object model ([0039] via “The local surface normal vectors are used both in collision avoidance calculations (where the vectors are used to calculate an amount of penetration of the gripper into the finger, and vice versa, where this penetration or interference is penalized in a constraint function) ….”), ([0040] via “In box 430, the grasp searching problem is modeled as an optimization, and one iteration is computed. To compute stable grasps, surface contacts and rigorous mathematic quality are adopted in the modeling. Gripper-object collision is also penalized in the optimization to avoid penetration, as mentioned above.”).
Further, Petersen teaches minimizing self-collision of the robot hand model ([0010] via “In another aspect, determining, within the first grasp region, a grasp, comprises determining the grasp based, at least in part, on a collision constraint associated with the robotic device when manipulating the object. In another aspect, the collision constraint is a self-collision constraint of the robotic device and/or an external collision constraint of the robotic device with an object in an environment of the robotic device.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Fan ‘599 wherein searching within the subspace of the configuration space of the robot hand model for the plurality of robot hand configurations comprises minimizing penetration of the robot hand model into the object model. Doing so outputs a stable grasp of the object model when penetrations are avoided, as stated above by Fan ‘599 in paragraph [0040].
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Petersen wherein searching within the subspace of the configuration space of the robot hand model for the plurality of robot hand configurations comprises minimizing self-collision of the robot hand model. Doing so allows the robot hand model to be able to grasp the object without conflicting with itself in the form of the self-collisions, as stated by Petersen ([0107] via “In some embodiments, the joint optimization of the grasp on an object and the whole-body trajectory of the robot may be subject to various constraints including, but not limited to, a balance constraint of the robot when manipulating the object, a collision constraint (e.g., a self-collision constraint and/or an external collision constraint) associated with the robot when manipulating the object, ….”).
12. Claim(s) 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), and further in view of Christen et al. (US 20240157557 A1 hereinafter Christen).
Regarding Claim 7, modified reference Huang teaches the method of claim 6, but is silent on wherein generating the set of candidate grasps based on the robot hand configurations having grasp quality scores above the grasp quality score threshold comprises: populating a given candidate grasp with a post-grasp posture from a given robot hand configuration having a grasp quality score above the grasp quality score threshold; extracting a post-grasp posture from a given robot hand configuration having a grasp quality score above the grasp quality score threshold; determining a pre-grasp posture to associate with the post-grasp posture; and populating the given candidate grasp with the pre-grasp posture and the post-grasp posture.
However, Christen teaches populating a given candidate grasp with a post-grasp posture from a given robot hand configuration having a grasp quality score above the grasp quality score threshold; extracting a post-grasp posture from a given robot hand configuration having a grasp quality score above the grasp quality score threshold ([0086] via “The grasp predictor functionality 218 determines when the virtual end-effector 109V is in the pre-grasp position, which occurs at the end of the approaching phase 240. The grasp predictor functionality 218
may calculate whether to switch from the approaching phase 240 to a grasping phase 244. The grasping phase 244 may include three components: 1) moving the virtual end-effector
109V forward from the pre-grasp pose to a grasping pose; 2) causing the virtual end-effector
109V to grip the virtual object(s) 112V (e.g., by closing the virtual end-effector 109V); and 3) retracting the virtual end-effector 109V, which may be holding the virtual object(s) 112V. After the grasping phase 244 is complete, an episode may end. Thus, an episode may extend from the beginning of the approaching phase 240 (e.g., the initial pose) to the end the grasping phase 244 (e.g., the goal pose).”), ([0089] via “Each of at least one pre-grasp pose may be labeled by positioning the virtual device 104V in the pre-grasp pose in the virtual environment 108V provided by the physics simulation module 134, executing the grasping motion, and labeling the pre-grasp pose as being successful (e.g., assigning a value of one to the pre-grasp pose) when the grasp is successful, or labeling the pre-grasp pose as being unsuccessful (e.g., assigning a value of zero to the pre-grasp pose) otherwise.”), ([0132] via “The finetuning control policy functionality 216-FT may explore the virtual environment 108V and receive a reward value (e.g., a sparse reward) based on successfully grasping the virtual object(s) 112V. For example, a reward value may be assigned to each of the transitions 296
within a plan generated for a particular episode. By way of a non-limiting example, the reward value may be one if the task has been completed successfully, otherwise the reward value may be zero.”), (Note: The Examiner interprets the closed gripping/grasping of the object of Christen as the post-grasp posture of the robot hand and the successfulness of the grasp of Christen as the grasp quality score.);
determining a pre-grasp posture to associate with the post-grasp posture ([0086] via “The grasp predictor functionality 218 determines when the virtual end-effector 109V is in the pre-grasp position, which occurs at the end of the approaching phase 240. The grasp predictor functionality 218 may calculate whether to switch from the approaching phase 240
to a grasping phase 244. The grasping phase 244 may include three components: 1) moving the virtual end-effector 109V forward from the pre-grasp pose to a grasping pose; 2) causing the virtual end-effector 109V to grip the virtual object(s) 112V (e.g., by closing the virtual end-effector 109V); and 3) retracting the virtual end-effector 109V, which may be holding the virtual object(s) 112V. After the grasping phase 244 is complete, an episode may end. Thus, an episode may extend from the beginning of the approaching phase 240 (e.g., the initial pose) to the end the grasping phase 244 (e.g., the goal pose).”), (Note: The Examiner interprets the pre-grasp position of Christen as the pre-grasp posture.); and
populating the given candidate grasp with the pre-grasp posture and the post-grasp posture ([0086] via “The grasping phase 244 may include three components: 1) moving the virtual end-effector 109V forward from the pre-grasp pose to a grasping pose; 2) causing the virtual end-effector 109V to grip the virtual object(s) 112V (e.g., by closing the virtual end-effector 109V); and 3) retracting the virtual end-effector 109V, which may be holding the virtual object(s) 112V.”), ([0089] via “Each of at least one pre-grasp pose may be labeled by positioning the virtual device 104V in the pre-grasp pose in the virtual environment 108V provided by the physics simulation module 134, executing the grasping motion, and labeling the pre-grasp pose as being successful (e.g., assigning a value of one to the pre-grasp pose) when the grasp is successful, or labeling the pre-grasp pose as being unsuccessful (e.g., assigning a value of zero to the pre-grasp pose) otherwise. A function that expresses binary cross-entropy loss may be used as an object function during training and model parameters that minimize the object function may be selected and used by the machine learning process(es) of the grasp predictor functionality 218.”), ([0132] via “The grasp predictor functionality 218 may be used with respect to the expert control policy functionality 290
and/or the finetuning control policy functionality 216-FT to determine when the virtual device 104V has reached a pre-grasp pose and the approaching phase 240 has ended. The finetuning control policy functionality 216-FT may explore the virtual environment 108V and receive a reward value (e.g., a sparse reward) based on successfully grasping the virtual object(s) 112V. For example, a reward value may be assigned to each of the transitions 296
within a plan generated for a particular episode. By way of a non-limiting example, the reward value may be one if the task has been completed successfully, otherwise the reward value may be zero.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Christen wherein generating the set of candidate grasps based on the robot hand configurations having grasp quality scores above the grasp quality score threshold comprises: populating a given candidate grasp with a post-grasp posture from a given robot hand configuration having a grasp quality score above the grasp quality score threshold; extracting a post-grasp posture from a given robot hand configuration having a grasp quality score above the grasp quality score threshold; determining a pre-grasp posture to associate with the post-grasp posture; and populating the given candidate grasp with the pre-grasp posture and the post-grasp posture. Doing so determines entire grasping sequences for the robot that result in satisfying the grasp quality threshold, as stated above by Christen in paragraphs [0089] and [0132].
13. Claim(s) 8 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), further in view of Christen et al. (US 20240157557 A1 hereinafter Christen), and further in view of Rodriguez Garcia et al. (US 20210146532 A1 hereinafter Rodriguez Garcia) and Norton (US 20210178609 A1 hereinafter Norton).
Regarding Claim 8, modified reference Huang teaches the method of claim 7, but is silent on wherein determining the pre-grasp posture to associate with the post-grasp posture comprises: determining a closing motion to transform the robot hand model from the post-grasp posture to a closed hand posture that is more closed compared to the post-grasp posture; determining a trajectory to transform the robot hand model from the closed hand posture to an open hand posture that is more open compared to the pre-grasp posture based on the closing motion; and generating the pre-grasp posture based on the open hand posture.
However, Rodriguez Garcia teaches determining a closing motion to transform the robot hand model from the post-grasp posture to a closed hand posture that is more closed compared to the post-grasp posture ([0050] via “For example, the first end effector 120a may be commanded to increase the magnitude of the force it applies to the object 150 (indicated by the straight arrow in FIG. 4B), while the second end effector 120b may be commanded to rotate to adjust the direction of the force it applies to the object 150 (indicated by the curved arrow in FIG. 4B).”), ([0052] via “More generally, monitoring slippage may include detecting that a position and/or an orientation of an object relative to an end effector has changed. If slippage is detected, the contact state may be adjusted based on the monitored slippage at 536. The contact state may be adjusted by adjusting a magnitude of an applied force between the end effector and the object (as indicated at 537), or by adjusting a direction of the applied force (as indicated at 538), as described above.”), (Note: The Examiner interprets the requirement of adjusting the magnitude of applied force, such as increasing the applied force, of Rodriguez Garcia as satisfying this limitation.).
Further, Norton teaches determining a trajectory to transform the robot hand model from the closed hand posture to an open hand posture that is more open compared to the pre-grasp posture based on the closing motion ([0046] via “One or both of the fingers 30 are movable relative to the body 20 for the gripping device 10 to be movable between an open orientation (as illustrated in FIG. 1) and a closed orientation. In the open orientation, the fingers 30 are spaced apart to be positioned on opposing sides of the tool 100.”), (Note: The Examiner interprets the movement between the closed and open orientations of Norton as the trajectory.); and
generating the pre-grasp posture based on the open hand posture ([0046] via “In the open orientation, the fingers 30 are spaced apart to be positioned on opposing sides of the tool 100.”), (Note: The Examiner interprets the open orientation of Norton as the pre-grasp posture.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Rodriguez Garcia wherein determining the pre-grasp posture to associate with the post-grasp posture comprises: determining a closing motion to transform the robot hand model from the post-grasp posture to a closed hand posture that is more closed compared to the post-grasp posture. Doing so adjusts the grasp to better exert the contact state, as stated above by Rodriguez Garcia in paragraph [0052].
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Norton wherein determining the pre-grasp posture to associate with the post-grasp posture comprises: determining a trajectory to transform the robot hand model from the closed hand posture to an open hand posture that is more open compared to the pre-grasp posture based on the closing motion; and generating the pre-grasp posture based on the open hand posture. Doing so adjusts the configuration of the robot hand so that it is able to grasp an item, as stated by Norton ([0050] via “In one example as illustrated in FIG. 1, each finger 30 is movable relative to the body 20 to adjust the grip 40 between the open and closed orientations. Each of the fingers 30 can include the same amount of movement. As illustrated in FIG. 1, each finger 30
moves the same amount in an inward direction indicated by arrows A and an outward direction indicated by arrows B.”), and shown in Figure 1 of Norton.
14. Claim(s) 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), further in view of Christen et al. (US 20240157557 A1 hereinafter Christen), further in view of Rodriguez Garcia et al. (US 20210146532 A1 hereinafter Rodriguez Garcia) and Norton (US 20210178609 A1 hereinafter Norton), and further in view of Bennice et al. (US 12168296 B1 hereinafter Bennice).
Regarding Claim 9, modified reference Huang teaches the method of claim 8, but is silent on wherein determining the closing motion comprises: determining a closed position within a finger volume formed by the post-grasp posture; assigning a set of virtual fingers to the robot hand model for the grasp type; and determining the closing motion to transform the set of virtual fingers from the post-grasp posture to the closed hand posture at the closed position.
However, Bennice teaches determining a closed position within a finger volume formed by the post-grasp posture; assigning a set of virtual fingers to the robot hand model for the grasp type; and determining the closing motion to transform the set of virtual fingers from the post-grasp posture to the closed hand posture at the closed position (Col. 17 lines 48-59, where “At t14, the user issues command 518A directing the simulated robot to move to the third waypoint which may, for example, represent a point where a gripper of the simulated robot may be situated around a simulated object to be picked up. Between t17 and t18, the user issues command 520A directing the simulated robot to close the gripper which may, for example, cause the gripper of the simulated robot to close around the simulated object. At t20 the user issues command 522A directing the simulated robot to move to the fourth waypoint which may, for example, correspond to a bin in which the simulated robot is to place the simulated object.”), (Note: See Figures 5A-C of Bennice as well.).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Bennice wherein determining the closing motion comprises: determining a closed position within a finger volume formed by the post-grasp posture; assigning a set of virtual fingers to the robot hand model for the grasp type; and determining the closing motion to transform the set of virtual fingers from the post-grasp posture to the closed hand posture at the closed position. Doing so simulates a full grasping motion of the robot hand to grasp an object, as stated above by Bennice.
15. Claim(s) 10 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), further in view of Christen et al. (US 20240157557 A1 hereinafter Christen), and further in view of Bennice et al. (US 12168296 B1 hereinafter Bennice).
Regarding Claim 10, modified reference Huang teaches the method of claim 7, but is silent on further comprising determining a grasp trajectory to transform the robot hand model between the pre-grasp posture and the post-grasp posture for the given candidate grasp and populating the given candidate grasp with the grasp trajectory.
However, Bennice teaches determining a grasp trajectory to transform the robot hand model between the pre-grasp posture and the post-grasp posture for the given candidate grasp (Col. 17 lines 46-59, where “Between t11 and t12, the user issues command 516A directing the simulated robot to open its gripper. At t14, the user issues command 518A directing the simulated robot to move to the third waypoint which may, for example, represent a point where a gripper of the simulated robot may be situated around a simulated object to be picked up. Between t17 and t18, the user issues command 520A directing the simulated robot to close the gripper which may, for example, cause the gripper of the simulated robot to close around the simulated object. At t20 the user issues command 522A directing the simulated robot to move to the fourth waypoint which may, for example, correspond to a bin in which the simulated robot is to place the simulated object.”), (Note: See Figures 5A-C of Bennice as well.) and populating the given candidate grasp with the grasp trajectory (Col. 13 lines 26-29, where “Episode data is generated based on the simulated episodes in which one or more of the robot control policies are used to perform one or more simulated tasks and is stored, at least temporarily, in episode log 330.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Bennice wherein the method further comprises determining a grasp trajectory to transform the robot hand model between the pre-grasp posture and the post-grasp posture for the given candidate grasp and populating the given candidate grasp with the grasp trajectory. Doing so simulates a full grasping motion of the robot hand to grasp an object, as stated above by Bennice in Col. 17 lines 46-59.
16. Claim(s) 11 and 12 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), further in view of Fan et al. '543 (US 20240198543 A1 hereinafter Fan '543).
Regarding Claim 11, modified reference Huang teaches the method of claim 1, but is silent on the method further comprising: determining a stable object pose for the object model; and determining a hand pose for the robot hand model to grasp the object model at the stable object pose; wherein searching within the subspace of the configuration space of the robot hand model for the plurality of robot hand configurations to engage the target graspable part of the object model with the grasp type comprises positioning the robot hand model at the hand pose and positioning the object model at the stable object pose.
However, Fan ‘543 teaches determining a stable object pose for the object model ([0036] via “A shape model of the workpiece (e.g., a CAD solid model) is provided at box 342, as is a model of the bin …. Using the workpiece and bin models, along with a defined number of workpieces to include in each bin, the bin containing many workpieces in random poses can be computed at the box 340.”), ([0037] via “The bin picking simulation at the box 350
recursively matches a grasp … with a pose of a particular workpiece in the bin, ….”), (Note: The Examiner interprets the stable object pose as a resting pose of an object on a ground surface, as the stable object pose is described within paragraph [0125] of the specification of the instant application.); and
determining a hand pose for the robot hand model to grasp the object model at the stable object pose ([0043] via “A specified design of a gripper 530 with fingertips 540 is used in a simulation to pick and remove one workpiece at a time until either no workpieces remain in the bin 510 or any remaining workpieces are not pickable due to collisions (such as the gripper body with a bin wall). A coordinate frame is shown on the gripper 530 to illustrate that the gripper 530 is placed in a particular position and orientation in order to cause the fingertips 540 to match the location of a suitable grasp on one of the workpieces 520.”);
wherein searching within the subspace of the configuration space of the robot hand model for the plurality of robot hand configurations to engage the target graspable part of the object model with the grasp type comprises positioning the robot hand model at the hand pose and positioning the object model at the stable object pose ([0037] via “The bin picking simulation at the box 350 recursively matches a grasp … with a pose of a particular workpiece in the bin, ….”), ([0043] via “As should be understood from the preceding discussion of FIG. 3, the gripping and collision detection simulation at the box 350 includes simulating a bin 510
with a plurality of workpieces 520 in a random pile or jumble. A specified design of a gripper
530 with fingertips 540 is used in a simulation to pick and remove one workpiece at a time until either no workpieces remain in the bin 510 or any remaining workpieces are not pickable due to collisions (such as the gripper body with a bin wall). A coordinate frame is shown on the gripper 530 to illustrate that the gripper 530 is placed in a particular position and orientation in order to cause the fingertips 540 to match the location of a suitable grasp on one of the workpieces 520.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Fan ‘543 wherein the method further comprises: determining a stable object pose for the object model; and determining a hand pose for the robot hand model to grasp the object model at the stable object pose; wherein searching within the subspace of the configuration space of the robot hand model for the plurality of robot hand configurations to engage the target graspable part of the object model with the grasp type comprises positioning the robot hand model at the hand pose and positioning the object model at the stable object pose. Doing so optimizes the grasping of objects by selecting a grasp that best matches the current poses of both the robot hand and the object, as state above by Fan ‘543 in paragraph [0043].
Regarding Claim 12, modified reference Huang teaches the method of claim 11, but is silent on wherein executing the simulated grasp in the physics engine to cause the robot hand model to engage the object model with the simulated grasp comprises positioning the robot hand model at the hand pose and positioning the object model at the stable object pose.
However, Fan ‘543 teaches wherein executing the simulated grasp in the physics engine to cause the robot hand model to engage the object model with the simulated grasp comprises positioning the robot hand model at the hand pose and positioning the object model at the stable object pose ([0037] via “The bin picking simulation at the box 350 recursively matches a grasp … with a pose of a particular workpiece in the bin, ….”), ([0043] via “As should be understood from the preceding discussion of FIG. 3, the gripping and collision detection simulation at the box 350 includes simulating a bin 510 with a plurality of workpieces 520 in a random pile or jumble. A specified design of a gripper 530 with fingertips 540 is used in a simulation to pick and remove one workpiece at a time until either no workpieces remain in the bin 510 or any remaining workpieces are not pickable due to collisions (such as the gripper body with a bin wall). A coordinate frame is shown on the gripper 530 to illustrate that the gripper 530 is placed in a particular position and orientation in order to cause the fingertips 540 to match the location of a suitable grasp on one of the workpieces 520.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Fan ‘543 wherein executing the simulated grasp in the physics engine to cause the robot hand model to engage the object model with the simulated grasp comprises positioning the robot hand model at the hand pose and positioning the object model at the stable object pose. Doing so optimizes the grasping of objects by selecting a grasp that best matches the current poses of both the robot hand and the object, as stated above by Fan ‘543 in paragraph [0043].
17. Claim(s) 13, 14, and 15 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), further in view of Li et al. (US 20190001508 A1 hereinafter Li).
Regarding Claim 13, modified reference Huang teaches the method of claim 1, but is silent on wherein generating the simulated grasp based on the given candidate grasp comprises adjusting a value of at least one parameter of the simulated grasp to be different from a value of a corresponding parameter of the given candidate grasp.
However, Li teaches wherein generating the simulated grasp based on the given candidate grasp comprises adjusting a value of at least one parameter of the simulated grasp to be different from a value of a corresponding parameter of the given candidate grasp ([0143] via “According to this configuration, …, a finger corresponding to a contact position whose contribution to the grasp stability is low and can be changed is determined as the finger-to-be-moved. When the grasp quality contribution degree Q.sub.10 of the finger-to-be-moved (e.g., the finger 10) is lower than the grasp quality contribution degrees … of other fingers, …, or when the contact force of the finger-to-be-moved 10 supplied to the object O decreases, it is less likely to drop the object due to a considerable decrease in the grasp stability of the object O. Further, by moving the finger-to-be-moved 10, for example, to a contact position at which the grasp quality contribution degree Q.sub.10 increases, it is possible to improve the grasp stability of the object O.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Li wherein generating the simulated grasp based on the given candidate grasp comprises adjusting a value of at least one parameter of the simulated grasp to be different from a value of a corresponding parameter of the given candidate grasp. Doing so iteratively adjusts the initial grasp to improve the grasp of the object by the robot hand, as stated above by Li.
Regarding Claim 14, modified reference Huang teaches the method of claim 13, but is silent on wherein the at least one parameter describes a post-grasp posture for the robot hand model to grasp the object model.
However, Li teaches wherein the at least one parameter describes a post-grasp posture for the robot hand model to grasp the object model ([0143] via “According to this configuration, …, a finger corresponding to a contact position whose contribution to the grasp stability is low and can be changed is determined as the finger-to-be-moved. When the grasp quality contribution degree Q.sub.10 of the finger-to-be-moved (e.g., the finger 10) is lower than the grasp quality contribution degrees … of other fingers, …, or when the contact force of the finger-to-be-moved 10 supplied to the object O decreases, it is less likely to drop the object due to a considerable decrease in the grasp stability of the object O. Further, by moving the finger-to-be-moved 10, for example, to a contact position at which the grasp quality contribution degree Q.sub.10 increases, it is possible to improve the grasp stability of the object O.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Li wherein the at least one parameter describes a post-grasp posture for the robot hand model to grasp the object model. Doing so iteratively adjusts the initial grasp to improve the grasp of the object by the robot hand, as stated above by Li.
Regarding Claim 15, modified reference Huang teaches the method of claim 13, but is silent on wherein adjusting the value of at least one parameter of the simulated grasp to be different from the value of the corresponding parameter of the given candidate grasp comprises adjusting at least one degree of freedom of the robot hand model to increase a degree of closeness of at least one robotic finger of the robot hand model to the object model.
However, Li teaches wherein adjusting the value of at least one parameter of the simulated grasp to be different from the value of the corresponding parameter of the given candidate grasp comprises adjusting at least one degree of freedom of the robot hand model to increase a degree of closeness of at least one robotic finger of the robot hand model to the object model ([0143] via “According to this configuration, …, a finger corresponding to a contact position whose contribution to the grasp stability is low and can be changed is determined as the finger-to-be-moved. When the grasp quality contribution degree Q.sub.10 of the finger-to-be-moved (e.g., the finger 10) is lower than the grasp quality contribution degrees … of other fingers, …, or when the contact force of the finger-to-be-moved 10
supplied to the object O decreases, it is less likely to drop the object due to a considerable decrease in the grasp stability of the object O. Further, by moving the finger-to-be-moved 10, for example, to a contact position at which the grasp quality contribution degree Q.sub.10 increases, it is possible to improve the grasp stability of the object O.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Li wherein adjusting the value of at least one parameter of the simulated grasp to be different from the value of the corresponding parameter of the given candidate grasp comprises adjusting at least one degree of freedom of the robot hand model to increase a degree of closeness of at least one robotic finger of the robot hand model to the object model. Doing so iteratively adjusts the initial grasp to improve the grasp of the object by the robot hand, as stated above by Li.
18. Claim(s) 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), further in view of Nagarajan et al. (US 10981272 B1 hereinafter Nagarajan).
Regarding Claim 16, modified reference Huang teaches the method of claim 1, further comprising: repeating simulating grasping of the object model with the robot hand model a plurality of times for a given candidate grasp for a plurality of candidate grasps from the set of candidate grasps, wherein each repeating simulating grasping of the object model with the robot hand model generates a plurality of simulated grasps with assigned grasp stability scores ([0060] via “The physics simulator 110 simulates effects of the various grasping models 116
on the various deformable object models 104 to generate grasp performance metrics 126 that are associated in the storage system 102 with a corresponding one of the deformable object models 104 and grasping models 116. The stored grasp performance metrics 112 and robotic grasp features 118 are input to a classifier 106 that generates a measure of correlation between them.”), ([0075] via “In block 204, grasp performance metrics are obtained from the simulation for stress, deformation, controllability, and instability of the response to the grasping model. … In block 208, it is determined whether a conforming model (of the grasping models) comprises robotic grasp features that meet manipulation constraints of the deformable object. In block 210, if (on condition that) the robotic grasp features from the simulation meet the manipulation constraints, a path planner for the robotic manipulator is configured with grasp settings associated with the grasping model.”).
Huang is silent on generating a machine learning training dataset based at least in part on the plurality of simulated grasps with assigned grasp stability scores.
However, Nagarajan teaches generating a machine learning training dataset based at least in part on the plurality of simulated grasps with assigned grasp stability scores (Col. 5 lines 41-65, where “Robots 110 may send data describing grasp attempts and their outcomes to the server system 120, which can collect the data and use it to improve the performance of the grasp model. For example, the server system 120 may annotate and analyze the data, and update the neural network grasp model by training the neural network grasp model on the data. … After updating the neural network model, the server system 120 may distribute the updated grasp model to one or more robots 110 in the fleet, effectively allowing all the robots 110 to learn improved grasping behavior from the collective experience of the fleet.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Nagarajan wherein the method further comprises generating a machine learning training dataset based at least in part on the plurality of simulated grasps with assigned grasp stability scores. Doing so improves the way the robot(s) grasps the object by learning from unsuccessful grasps why the grasp was unsuccessful and how to improve it, as stated above by Nagarajan.
19. Claim(s) 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Huang et al. (US 20220297297 A1 hereinafter Huang) in view of Ciocarlie et al. ("Dimensionality reduction for hand-independent dexterous robotic grasping" hereinafter Ciocarlie) and Fan et al. '602 (US 20230256602 A1 hereinafter Fan '602), further in view of Jin et al. (US 20210030483 A1 hereinafter Jin) and Edsinger (US 10275543 B1 hereinafter Edsinger).
Regarding Claim 17, modified reference Huang teaches the method of claim 1, but is silent on the method further comprising: extracting a kinematic description of a robotic hand from a robot model; and generating the robot hand model based at least in part on the kinematic description of the robotic hand.
However, Jin teaches extracting a kinematic description of a robotic hand from a robot model ([0035] via “The robotic surgery device 114 can include a nominal position estimator module 118 to estimate the nominal position of the surgery tool 126 based on the relative positioning of the camera 128 and the surgery tool 126 as determined based on the robotic arm kinematics. The prediction can be performed through a kinematic chain model of the robotic arm 150 of the robotic surgery device 114 holding the surgery tool 126 and a kinematic chain model of the robotic arm 150D holding the camera 128. The kinematic chain model can estimate the nominal position of the surgery tool 126 based on the dimensions and connectivity of the links and joints of the robotic arm 150. … In addition, computer aided design (CAD) models of the surgery tool can also be used to estimate the nominal position of the surgery tool 126.”).
Further, Edsinger teaches generating the robot hand model based at least in part on the kinematic description of the robotic hand (Col. 2 lines 18-36, where “Upon detecting that a tool database includes a model of one or more fingertips for the given object, the operations may also include providing the model of the one or more one or more fingertips for the given object from the tool database. Upon detecting that the tool database excludes the model of the one or more fingertips for grasping the given object, the operations may involve determining a model of one or more fingertips for the given object. Determining the model of one or more fingertips may involve iterating over a plurality of fingertip geometries to determine one or more particular fingertip geometries that are compliant to a shape of the given object at a grasp region on the given object, iterating the one or more particular fingertip geometries over a plurality of fingertip sizes to determine one or more particular fingertip sizes that are compliant to one or more dimensions of the given object at the grasp region of the given object, and providing the determined model of the one or more fingertips having the one or more particular fingertip geometries and the one or more particular fingertip sizes.”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Jin wherein the method further comprises extracting a kinematic description of a robotic hand from a robot model. Doing so uses properties of the robot to determine its relative location within the space and relative to itself, as stated above by Jin.
In addition, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teachings of Edsinger wherein the method further comprises generating the robot hand model based at least in part on the kinematic description of the robotic hand. Doing so configures a robot hand to grasp a new object, even if the new object grasp has not been previously taught to the robot, as stated above by Edsinger.
Examiner’s Note
20. The Examiner has cited particular paragraphs or columns and line numbers in the
references applied to the claims above for the convenience of the Applicant. Although the
specified citations are representative of the teachings of the art and are applied to specific
limitations within the individual claim, other passages and figures may apply as well. It is
respectfully requested of the Applicant in preparing responses, to fully consider the references
in their entirety as potentially teaching all or part of the claimed invention, as well as the
context of the passage as taught by the prior art or disclosed by the Examiner. See MPEP
2141.02 [R-07.2015] VI. A prior art reference must be considered in its entirety, i.e., as a whole,
including portions that would lead away from the claimed Invention. W.L. Gore & Associates,
Inc. v. Garlock, Inc., 721 F.2d 1540, 220 USPQ 303 (Fed. Cir. 1983), cert, denied, 469 U.S. 851
(1984). See also MPEP §2123.
Conclusion
21. Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
22. Any inquiry concerning this communication or earlier communications from the
examiner should be directed to BYRON X KASPER whose telephone number is (571)272-3895.
The examiner can normally be reached Monday - Friday 8 am - 5 pm EST.
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, Adam Mott can be reached on (571) 270-5376. 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.
/BYRON XAVIER KASPER/Examiner, Art Unit 3657
/ADAM R MOTT/Supervisory Patent Examiner, Art Unit 3657
