Natural Language Agent Mapping Engineering Drawings to Manufacturers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for matching manufacturing supplier capabilities to specific requirements are inflexible, inaccurate, and time-consuming, particularly for small-to-medium manufacturers, due to disparate data sources, limited training, and complexity.
Innovation Solution
A system that uses natural language agent programs and large language models to process user input and extracted data from engineering drawings, generating a list of suitable manufacturers by mapping manufacturing requirements to qualified suppliers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If conventional rules-based logic or form-driven techniques are used to match buyer requirements to manufacturer capabilities, then the matching process can be automated to some extent, but the results are inflexible, inaccurate, and time-consuming due to disparate data sources and limited training
Solution Approach 1:
The patent replaces conventional rules-based logic and form-driven techniques with a large language model (LLM) system that processes natural language queries and engineering drawing data. This substitution enables the system to understand and match manufacturing requirements to manufacturer capabilities with high accuracy by leveraging the LLM's ability to comprehend complex, unstructured data and perform semantic matching, thereby resolving the contradiction between automation extent and matching precision
Solution Approach 2:
The system transforms the matching approach by changing from rigid rule-based parameters to flexible natural language processing parameters. The LLM analyzes engineering drawings, extracts manufacturing requirements, and compares them against manufacturer capabilities using semantic understanding rather than predefined rules, enabling accurate matching while maintaining full automation
2Measurement precision
If manual searching and evaluation of manufacturing suppliers is performed, then accurate matching of requirements to capabilities can be achieved, but the process becomes time-consuming and expensive requiring skilled personnel
Solution Approach 1:
The system enables self-service by allowing buyers to input their manufacturing requirements in natural language along with engineering drawings, and the LLM automatically performs the entire matching process without requiring skilled procurement personnel. The system independently extracts requirements, identifies suitable manufacturers, and returns matched results, eliminating the need for manual searching while maintaining high accuracy
Solution Approach 2:
The large language model acts as an intermediary between the buyer's requirements and the manufacturer database. It processes natural language inputs, interprets engineering drawing data, and automatically matches requirements to manufacturer capabilities, replacing the need for skilled human intermediaries while maintaining matching accuracy and reducing time loss
3Manufacturing precision
If 2D drawing packages are created from 3D CAD models by skilled draftspersons, then accurate manufacturing specifications can be produced, but the process incurs significant cost and time
Solution Approach 1:
The patent replaces the mechanical process of manual drafting by skilled draftspersons with an automated LLM-based system that directly processes 3D CAD models and engineering drawings. The system extracts manufacturing specifications automatically using computer vision and natural language processing, eliminating the need for human draftspersons while maintaining specification accuracy and dramatically improving productivity
4Ease of manufacture
If small-to-medium manufacturers have few resources to publicize their capabilities, then cost structures remain simple, but finding these manufacturers becomes difficult for buyers
Solution Approach 1:
The LLM-based matching system serves as an intelligent intermediary that actively searches and matches manufacturers to buyer requirements. Instead of requiring manufacturers to publicize their capabilities extensively, the system queries the manufacturer database using extracted requirements and identifies suitable matches, thereby easing the burden on small-to-medium manufacturers while improving ease of locating suitable manufacturers for buyers
Data Source
AI summary
Techniques for computer science, data science, data analytics, computer software, algorithmic analysis, and networked technologies for sourcing, procurement, manufacturing, and supply chain management in small-to-medium manufacturing (“SMM”) industries that involve processes whereby implementing user input and extracted image and text from engineering drawings may be used to map manufacturing requirements to a subset of manufacturers via natural language agent programs. More specifically, natural language agent programs may be implemented to apply input and extracted image and text from engineering drawings to a large language model (“LLM”). An example method may include receiving a user input including data representing requirements to manufacture a physical structure, concatenating an engineering drawing summary data, a shoptype description, and estimated part size instruction data to combine the at least two of the engineering drawing summary data, the shoptype description, and the estimated part size instruction data to generate a subset (e.g., a list) of qualified manufacturers (e.g., SMMs).