AI-Generated 3D Model Reconstruction for Faster Design Iteration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The current 3D CAD geometry design process is inefficient, requiring multiple sketches per perspective, extensive manual adjustments, and specialized expertise, with significant communication overhead and time-consuming iterations between designers and CAD professionals.

Innovation Solution

A method using trained artificial neural networks, particularly generative adversarial networks (GANs), to generate 3D models from 2D sketches by specifying an object type, training the network with base image data, and applying interpolation and recombination of image properties to meet design criteria, followed by reconstruction into a digital 3D model.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If 2D sketches are used to create 3D models, then the design process can start with simple 2D drawings, but multiple sketches per perspective are necessary and any design change requires individual adjustments to each sketch

Engineering Contradiction:
Improveease of creating initial designVSAvoidtime for adjustments
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent uses photogrammetry to capture multiple 2D images of the object from different angles and automatically generates a 3D model, eliminating the need to manually create multiple 2D sketches. The system copies real-world visual information and transforms it directly into 3D geometry, reducing manual adjustment time.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces the manual mechanical process of drawing and adjusting 2D sketches with an automated computational system. The photogrammetry software automatically processes multiple 2D images and reconstructs the 3D model through algorithmic operations, substituting human manual adjustment with automated image processing.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Shape

If 3D models are created based on 2D sketches, then a visual representation is obtained, but a media break occurs and specialized expertise is required

Engineering Contradiction:
Improve3D model visualizationVSAvoidspecialized expertise required
Core Design Contradiction:
ShapeVSDevice complexity

Solution Approach 1:

The patent enables non-expert users to create 3D models by capturing photos with standard cameras or smartphones. The photogrammetry software automatically performs the complex tasks of feature detection, matching, and 3D reconstruction without requiring user expertise in CAD or 3D modeling techniques.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent introduces photogrammetry software as an intermediary between simple 2D photo capture and 3D model creation. This intermediary system handles the complex transformation process, allowing users to work with familiar 2D images while the software manages the specialized 3D reconstruction operations.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Adaptability or versatility

If incremental design changes are made to 3D models, then design refinement is possible, but time-consuming manual adjustments and extensive communication between designers and CAD professionals are essential

Engineering Contradiction:
Improvedesign modification capabilityVSAvoidcommunication and adjustment time
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent creates a dynamic workflow where 3D models can be rapidly regenerated from updated 2D photos. Designers can capture new images reflecting design changes and the photogrammetry system automatically regenerates the 3D model, enabling flexible design iteration without manual CAD editing or extensive communication cycles.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4555485B1Designing device, and method for providing a 3D model for a design development process of an object
Publication Date: 2026.04.08 AUDI AG
  • EP4555485B1 patent drawingFigure 1
  • EP4555485B1 patent drawingFigure 2

AI summary

The invention relates to a designing device (1) and to a method for providing a 3D model (5) for a design development process of an object. The method comprises: predefining (S2) an object type (18) to a trained artificial neural network (10), the artificial neural network (10) being trained with a plurality of items of basic image data (13) of objects of said object type (18); generating (S3) an image data set (15) of the predefined object type using the trained artificial neural network (10), the artificial neural network (10) providing items of image data (14) from a plurality of perspectives as the image data set (15); checking (S4) the generated image data set (15) for the presence of a design selection criterion, items of image data (14) being regenerated by changing an image property criterion (S5) if the design selection criterion is not present (17); using a reconstruction algorithm to generate (S6) a digital 3D model (5) from the image data set (15) for which the design selection criterion is present (16); and providing (S7) the generated digital 3D model (5) as a design development process for the object.