Machine Learning Prioritization of Common-Asset Work Requests
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for prioritizing work orders on operational assets are often based on subjective decision-making, leading to sub-optimal scheduling, especially when multiple work orders are interlinked, and there is a need for a more efficient and objective approach to prioritize and schedule work requests.
Innovation Solution
A method and system using machine learning to prioritize work requests by encoding them as multidimensional representations, reducing dimensionality to one-dimensional structures while preserving variance factors, and utilizing a trained model to determine optimal sequences for performing work activities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If work requests are prioritized using subjective decision-making, then flexibility in scheduling is maintained, but prioritization accuracy and objectivity deteriorate
Solution Approach 1:
The patent replaces subjective human decision-making with an automated machine learning system that processes work request data objectively. The ML model analyzes multiple features (asset criticality, work urgency, resource availability) to generate prioritization scores, eliminating bias and inconsistency in manual prioritization while maintaining system flexibility through configurable parameters.
Solution Approach 2:
The patent introduces a machine learning model as an intermediary between work request submission and scheduling decisions. This intermediary processes multidimensional work request data, reduces it to essential features, and outputs prioritization recommendations that balance objectivity with operational flexibility, resolving the contradiction between accurate prioritization and system complexity.
2Measurement precision
If multidimensional work request data is processed in full detail, then prioritization accuracy is improved, but processing time and computational resources increase
Solution Approach 1:
The patent extracts only the most critical features from multidimensional work request data using dimensionality reduction techniques. The ML model identifies and retains key variables (asset criticality score, work urgency level, resource constraints) while discarding redundant information, thereby maintaining prioritization accuracy while significantly reducing processing time and computational burden.
Solution Approach 2:
The patent segments work request data into distinct dimensional categories (asset characteristics, work parameters, resource availability) and processes each segment separately through the ML model. This segmentation allows efficient handling of complex multidimensional data by breaking it into manageable feature groups, reducing overall processing time while preserving essential information for accurate prioritization.
Data Source
AI summary
Prioritizing a work request pertaining to a physical asset can include generating a data structure that encodes the work request as a multidimensional representation indicating at least one classification, each at least one classification including at least one sub-category. In response to identifying multiple work requests encoded as multidimensional representations with respect to the physical asset, each multidimensional representation can be reduced to a one-dimensional (1-D) representation that preserves a variance factor of each sub-category of each multidimensional representation. Each 1-D structure can be input to a machine learning model trained to prioritize each of the work requests. The work requests can be prioritized in accordance with the machine learning model based on the 1-D structures. The priorities of each of the work requests can be output.