Machine-Learning Design Feedback for Manufacturable Objects

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional object design methods lack real-time feedback on production costs and quality consequences, leading to non-ideal designs in terms of producibility and cost, due to static and limited manufacturing feedback, and outdated quality specifications.

Innovation Solution

A system utilizing a machine-learning model that analyzes design data and production data to provide live updates on design modifications, process conditions, and environmental factors, enabling real-time optimization of object design for reduced production costs and defect rates.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional design methods are used focusing on performance metrics, then design performance is improved, but production cost and manufacturability deteriorate

Engineering Contradiction:
Improvedesign performanceVSAvoidproducibility
Core Design Contradiction:
ReliabilityVSEase of manufacture

Solution Approach 1:

The patent implements a feedback mechanism where manufacturing feedback is provided to designers in real-time during the design process. This feedback loop allows designers to see the impact of their design choices on production cost and manufacturability, enabling them to adjust designs before finalization, thus resolving the contradiction between performance optimization and manufacturing ease

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent introduces an intermediary system (manufacturing feedback mechanism) that translates design decisions into manufacturing implications. This intermediary provides designers with actionable insights about production cost and manufacturability, bridging the gap between design performance and manufacturing ease without requiring designers to directly understand complex manufacturing processes

Inventive Principle:
Principle #24Intermediary (Mediator)

2Device complexity

If static manufacturing feedback is provided to designers, then design review is simplified, but design optimization for production cost deteriorates

Engineering Contradiction:
Improvefeedback system complexityVSAvoidproduction cost
Core Design Contradiction:
Device complexityVSEase of manufacture

Solution Approach 1:

The patent transforms static manufacturing feedback into a dynamic system that updates in real-time as designers make modifications. The feedback adapts to each design iteration, providing current production cost estimates and manufacturability assessments, thus enabling continuous optimization without proportionally increasing system complexity

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent provides preliminary manufacturing feedback early in the design process, allowing designers to make informed decisions before designs are finalized. This preliminary action prevents costly redesigns later and optimizes production cost from the outset, rather than addressing issues after design completion

Inventive Principle:
Principle #10Preliminary action

3Device complexity

If quality specifications are updated on a slow schedule, then document management is simplified, but production quality deteriorates

Engineering Contradiction:
Improvedocument update complexityVSAvoidproduction quality
Core Design Contradiction:
Device complexityVSManufacturing precision

Solution Approach 1:

The patent implements continuous updating of quality specifications rather than periodic updates. The system continuously monitors and updates quality requirements, ensuring designers always work with current specifications. This continuous action maintains high production quality without requiring complex manual document management processes

Inventive Principle:
Principle #20Continuity of useful action

Data Source

PatentEP3702982B1Object design using machine-learning model
Publication Date: 2023.10.04 THE BOEING CO
  • EP3702982B1 patent drawingFigure 1
  • EP3702982B1 patent drawingFigure 2
  • EP3702982B1 patent drawingFigure 3

AI summary

A system to aid in design for manufacturing an object includes a processor and a memory configured to store instructions. The processor is configured to receive first data representing a design of the object to be manufactured and second data representing a machine-learning model. The processor is configured to execute the instructions to generate third data using the first data and the second data. The third data indicates at least one of a modification to the design of the object or process conditions for production of the object. The processor is configured to send the design of the object, the process conditions, or both, to a manufacturing tool to enable production of the object. The machine-learning model is representative of production data and based at least partially on one or more of: object features, process parameters, environmental factors, and quality data.