AR Industrial Space Modeling Through Real-Time Equipment Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital modeling of industrial spaces with mixed equipment from different manufacturers is challenging due to the lack of numerical definitions for machines, leading to time-consuming and costly computer-designing processes, and existing methods like point cloud scanning fail to distinguish between machines and equipment, limiting reorganization possibilities.
Innovation Solution
A method involving three-dimensional digitization of the industrial space followed by real-time segmentation in a virtual environment using an encompassing shape to define virtual objects, allowing users to segment and configure equipment without detailed redrawing, using photogrammetry or lasergrammetry for scanning and virtual reality for precise segmentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a three-dimensional scan is performed to obtain a point cloud representing the production space, then the entire space including building, machines, and equipment can be captured, but the point cloud does not distinguish between building, machines, and equipment, making it useless for reorganizing production lines
Solution Approach 1:
The patent applies segmentation by dividing the point cloud data into distinct equipment groups using encompassing shapes. Each encompassing shape (convex hull) segments a specific piece of equipment from the overall point cloud, creating separate virtual objects for each machine while preserving the complete spatial representation. This resolves the contradiction by maintaining both completeness and distinguishability.
2Productivity
If a rough representation of each machine is created as blocks to mitigate modeling costs, then modeling time and cost are reduced, but it is impossible for the user to immediately distinguish one machine from another, and numerical modeling offers little added value
Solution Approach 1:
The patent uses copying by creating virtual objects that are digital representations of physical equipment. Instead of manually modeling each machine with detailed geometry, the system automatically generates simplified virtual objects (encompassing shapes) that copy the essential spatial boundaries and positions of equipment from the point cloud, providing both speed and distinguishability.
3Measurement precision
If detailed computer-generated models are created for each machine without numerical definitions to achieve accurate representation, then visual accuracy is improved, but the process becomes time-consuming and expensive
Solution Approach 1:
The patent applies preliminary action by performing three-dimensional scanning and point cloud generation automatically before the modeling phase. This preliminary capture of spatial data eliminates the need for time-consuming manual measurements and modeling, as the point cloud already contains accurate geometric information that can be directly converted into virtual objects.
4Ease of operation
If the designer defines the level of precision for modeling each machine without familiar knowledge of machine structure and function, then modeling can proceed without expert input, but the designer might draw a machine as a single block when it actually comprises several independent modules, reducing modeling accuracy
Solution Approach 1:
The patent applies self-service by enabling the end user to perform segmentation and virtual object creation themselves through the virtual reality interface. The user interacts with the system to define encompassing shapes around equipment they recognize, allowing them to control the segmentation based on their own knowledge of machine structures and functions, eliminating the need for expert designer intervention.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables users to efficiently and accurately model industrial spaces by allowing self-segmentation of equipment, reducing modeling time and costs, and providing a more realistic representation for reorganization, with improved precision and user control over equipment placement.
Implementation Method 1
using photogrammetry or lasergrammetry for scanning
Implementation Method 2
using photogrammetry or lasergrammetry for scanning
Data Source
Figure 1
Figure 2~3
Figure 4~6
AI summary
A digital modeling process for an industrial space, the industrial space comprising a plurality of equipment (E3) arranged in an industrial building (B), the modeling process comprising: a three-dimensional digitization step of the industrial space so as to obtain a sheet (NAP) defining meshes representing the external envelope of the industrial space with all of its equipment (E3), a real-time segmentation step of the sheet (NAP) in a virtual environment by a user during which: an encompassing shape (F) is positioned to encompass at least one piece of equipment to be segmented (E3), and the points of the sheet (NAP), belonging to the volume of the encompassing shape (F), are associated so as to form a virtual object (OBJ).