Hardware Component Identification Using Weighted Consensus Values
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aircraft hardware component suppliers and entities face inefficiencies in deriving consensus values due to unnecessary bandwidth usage, data entry errors, and long resolution times in existing processes, lacking access to consensus values and relying on large data lakes.
Innovation Solution
A machine-learned model is trained to determine consensus values for aircraft hardware components, utilizing a weighting model to apply weights to historical and new data, generating a searchable database with confidence outputs to accurately identify components with large variance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If large data lakes of historical training data are maintained to train the machine learning model, then the model can be re-trained based on new data, but the system complexity and data storage requirements increase significantly
Solution Approach 1:
The patent extracts only the essential training data needed for model re-training rather than maintaining complete data lakes. The system selectively retrieves and processes only relevant historical data and new data required for updating the model, eliminating the need to store and manage entire data lakes while preserving re-training capability.
2Adaptability or versatility
If existing processes use back-and-forth network communications to derive values, then multiple entities can exchange information, but bandwidth is wasted and resolution time increases
Solution Approach 1:
The patent introduces a centralized database as an intermediary that stores consensus values. Instead of multiple entities communicating back-and-forth, any entity can directly query the database to obtain the consensus value. This eliminates redundant network communications while maintaining multi-entity information exchange capability, significantly improving derivation efficiency.
3Adaptability or versatility
If manual data entry processes are used to collect hardware component values, then data can be gathered from multiple sources, but data entry errors increase reducing accuracy
Solution Approach 1:
The patent implements automated data collection processes where the system itself retrieves data from multiple sources without manual intervention. The machine learning model automatically processes incoming data, applies weighting algorithms, and updates consensus values. This self-service approach eliminates human data entry errors while maintaining the ability to collect data from multiple sources, significantly improving data accuracy.
4Measurement precision
If a searchable database is generated after extensive training and re-training iterations, then accurate consensus values can be provided, but the processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary training of the machine learning model using available historical data before the searchable database is needed. The model is pre-trained and the initial searchable database is generated in advance. When new data becomes available, the system performs efficient incremental updates rather than complete re-training, significantly reducing the time required to maintain accurate consensus values in the database.
Data Source
AI summary
A system and a method are disclosed for training a machine-learned model. A device retrieves entries from a database that each correspond a hardware component to a value. The device inputs the entries into a weighting model, and the weighting model outputs weights for the values. The device generates a training set including data formed by pairing each respective hardware component to its respective weighted value, and trains the machine-learned model using the training set. The device receives new data comprising a hardware component and a respective value, determines weights therefor, and re-trains the machine-learned model accordingly. Responsive to detecting a trigger, the device uses the machine-learned model to generate a searchable database, and outputs results to search queries including a value for a queried hardware component and a confidence that the value is correct.