Manufacturing assistance apparatus

Patent metadata

Field	Value
Publication number	US-12086734-B2
Application number	US-201816209743-A
Country	US
Kind code	B2
Filing date	Dec 4, 2018
Priority date	Jan 31, 2018
Publication date	Sep 10, 2024
Grant date	Sep 10, 2024

Abstract

A manufacturing assistance apparatus includes a learning unit and an estimator. The learning unit is configured to load a plurality of pieces of actual measurement data in each of which a gap and a plurality of parameters are associated with each other, and construct an estimation model on the basis of machine learning in which the plurality of pieces of actual measurement data serve as teacher data. The gap is provided between a first workpiece and a second workpiece that eventually structure an airframe of an aircraft and that are eventually fastened to each other. The estimation model estimates the gap from the plurality of parameters. The estimator is configured to derive an estimation value of a length of the gap on which measurement has not yet been performed, on the basis of the estimation model constructed by the learning unit and the plurality of parameters.

First claim

The invention claimed is:

1. A manufacturing assistance apparatus comprising at least one machine readable medium storing instructions and at least one processor configured to execute the instructions to:
load pieces of actual measurement data in each of which a gap and parameters are associated with each other;
construct a trained estimation model on a basis of machine learning in which the pieces of actual measurement data serve as teacher data, the gap being provided between a first workpiece and a second workpiece that eventually structure an airframe of an aircraft and that are eventually fastened to each other;
input the parameters into a trained latest estimation model to derive an estimation value of a length of the gap on which measurement has not yet been performed, wherein the trained latest estimation model is trained by inputting the teacher data comprising latest actual measurement data into the constructed trained estimation model to execute the constructed trained estimation model;
select, when determining that the estimation value of the length of the gap falls within an application range, a first standard shim from among predetermined standard shims;
identify an order of magnitude of influence by the parameters exerted on the estimation value of the length of the gap based on test data that is related to the parameters;
select a correction procedure and a second standard shim from among the predetermined standard shims, based on determining that (i) the estimation value of the length of the gap falls outside the application range, (ii) there is a selectable correction procedure from predetermined correction procedures to allow the gap to be closer to the application range, and (iii) the correction procedure has a largest magnitude of influence among the correction procedures and is selected based on the order of the magnitude of the influence by the parameters, each of the predetermined correction procedures being associated with one or more correction values for the parameters and stored in a storage, wherein the correction procedures correct the parameters to generate corrected input for the trained latest estimation model;
select, a dedicated shim different from the predetermined standard shims, based on determining that (i) the estimation value of the length of the gap falls outside the application range and (ii) there is no correction procedure selectable from the predetermined correction procedures to allow the gap to be closer to the application range;
update, in response to selecting the correction procedure, the corrected parameters into updated parameters corresponding to the selected correction procedure based on the one or more correction values, wherein
the predetermined correction procedures include replacing a related component part, performing a machining process on the related component part, changing positions of a first jig for the first workpiece or a second jig for the second workpiece, or changing a temperature around the related component part,
the related component part is the first workpiece, the second workpiece, the first jig for the first workpiece, or the second jig for the second workpiece,
the parameters include parameters for the first workpiece and parameters for the second workpiece,
the parameters for the first workpiece include: a thickness of a part of the first workpiece which is supported by the first jig; a thickness of a part of the first workpiece which is to be fastened to the second workpiece; a tilt of the first workpiece to a horizontal plane; or a height of a contact surface of the first jig which comes into contact with the first workpiece,
the parameters for the second workpiece include: a thickness of a part of the second workpiece which is to be fastened to the first workpiece; or a height of a contact surface of the second jig which comes into contact with the second workpiece, and
the at least one processor is configured to execute the instructions to:
in response to selecting the correction procedure, input the corrected parameters into the trained latest estimation model to derive a corrected estimation value of the length of the gap; and
select, where the corrected estimation value of the length of the gap falls within an application range, the second standard shim from among the predetermined standard shims.

2. The manufacturing assistance apparatus according to claim 1 , wherein the at least one processor is configured to execute the instructions to:
derive a median of lengths of gaps each of which corresponds to the gap from the pieces of the actual measurement data, and
determine whether any of the predetermined standard shims is available based on the estimation value of the length of the gap, the median and a tolerance of the gap.

3. The manufacturing assistance apparatus according to claim 1 , wherein the at least one processor is configured to execute the instructions to narrow down a tolerance of each of the parameters based on the order of the magnitude of the influence.

4. The manufacturing assistance apparatus according to claim 1 , wherein the parameters include positions of the first jig and the second jig, deformation amounts upon pressing the first workpiece against the first jig and the second workpiece against the second jig, and a tolerance measurement value after fastening the first workpiece with the second workpieces.

5. The manufacturing assistance apparatus according to claim 1 , wherein the at least one processor is configured to execute the instructions to:
repeat selecting a correction procedure from the predetermined correction procedures to be the correction procedure, in a case where there is still an applicable correction procedure that has not been selected in the predetermined correction procedures, until the standard shim is selected.

6. A manufacturing assistance apparatus comprising circuitry configured to:
load pieces of actual measurement data in each of which a gap and parameters are associated with each other, the gap being provided between a first workpiece and a second workpiece that eventually structure an airframe of an aircraft and that are eventually fastened to each other,
construct a trained estimation model on a basis of machine learning in which the pieces of actual measurement data serve as teacher data, the trained estimation model estimating the gap from the parameters,
input the parameters into the trained latest estimation model to derive an estimation value of a length of the gap on which measurement has not yet been performed, wherein the trained latest estimation model is trained by inputting the teacher data comprising latest actual measurement data into the constructed trained estimation model to execute the constructed trained estimation model,
select, when determining that the estimation value of the length of the gap falls within an application range, a first standard shim from among predetermined standard shims,
identify an order of magnitude of influence by the parameters exerted on the estimation value of the length of the gap based on test data that is related to the parameters,
select a correction procedure and a second standard shim of the predetermined standard shims, based on determining that (i) the estimation value of the length of the gap falls outside the application range, (ii) there is a selectable correction procedure from predetermined correction procedures to allow the gap to be closer to the application range, and (iii) the correction procedure has a largest magnitude of influence among the correction procedures and is selected based on the order of the magnitude of the influence by the parameters, each of the predetermined correction procedures being associated with one or more correction values for the parameters and stored in

Assignees

• Subaru Corp

Inventors

• Kobayashi Toshimasa
• Kikuchi Marosuke
• Sato Yoichi
• Matsuda Yuki
• Ogisu Toshimichi
• Hirabayashi Daisuke
• Suzuki Kei
• Onishi Toru
• Miyanishi Haruka

Classifications

• G05B19/41805 Primary

  characterised by assembly · CPC title

• G05B2219/45071

  Aircraft, airplane, ship cleaning manipulator, paint stripping · CPC title

• G06N20/00

  Machine learning · CPC title

• B64F5/10 Primary

  Manufacturing or assembling aircraft, e.g. jigs therefor · CPC title

• G06F17/11

  for solving equations {, e.g. nonlinear equations, general mathematical optimization problems (optimization specially adapted for a specific administrative, business or logistic context G06Q10/04)} · CPC title