Threshold-Based Maintenance for Series-Parallel Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial and mission-critical assets with series-parallel systems face challenges in optimal maintenance scheduling due to computationally intensive resource allocation, especially when resources are sparse and shared across multiple subsystems, leading to excessive downtime and costly breakdowns.
Innovation Solution
A threshold-based heuristic maintenance policy is implemented, which determines optimal resource allocation by balancing immediate and long-term rewards, prioritizing repairs based on maximum marginal reward and asset reliability, using a scalable linear backward recursion to derive near-optimal maintenance schedules in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If optimal maintenance scheduling is implemented using traditional resource allocation methods, then asset reliability is improved, but computational complexity and time required for resource allocation increases significantly
Solution Approach 1:
The patent segments the maintenance scheduling problem into discrete decision epochs and divides resources into allocable units. The series-parallel system is decomposed into independent subsystems with parallel components, allowing separate analysis and allocation decisions for each component group, thereby reducing overall computational complexity while maintaining reliability optimization.
Solution Approach 2:
The patent transforms the continuous resource allocation problem into a discrete parameter optimization problem by defining threshold-based maintenance triggers and integer resource allocation units. This parameter discretization enables the use of dynamic programming and heuristic algorithms that are computationally tractable while still achieving near-optimal asset reliability.
2Reliability
If traditional resource allocation methods are used for maintenance scheduling, then maintenance effectiveness is improved, but downtime increases due to excessive computational requirements
Solution Approach 1:
The patent pre-calculates threshold values and maintenance policies during idle periods or using historical data, storing these decisions for rapid deployment during operational periods. This preliminary action separates the computationally intensive optimization phase from the time-critical execution phase, reducing downtime during actual maintenance decisions.
Solution Approach 2:
The system uses real-time monitoring data and automated threshold comparisons to trigger maintenance actions without requiring complex real-time computational optimization. The pre-established policies enable the system to self-determine maintenance timing and resource allocation based on monitored component conditions, eliminating computational delays during critical periods.
3Reliability
If more maintenance resources are allocated to subsystems, then asset reliability is improved, but the cost of maintenance increases
Solution Approach 1:
The patent applies differentiated resource allocation strategies to different subsystems and components based on their individual reliability characteristics, criticality to asset operation, and failure consequences. High-criticality components receive prioritized resource allocation while less critical components receive minimal or preventive-only maintenance, optimizing the ratio of reliability improvement to resource consumption.
Solution Approach 2:
The patent implements threshold-based maintenance triggers that activate maintenance actions only when component degradation reaches predetermined levels, rather than applying continuous or excessive maintenance. This partial action approach maintains reliability by intervening only when necessary, avoiding unnecessary resource consumption on components that have not yet reached critical degradation states.
4Productivity
If real-time maintenance scheduling is implemented, then operational efficiency is improved, but computational resources and processing power are consumed
Solution Approach 1:
The system employs automated threshold monitoring and rule-based decision triggers that require minimal computational processing. Real-time monitoring data is automatically compared against pre-established thresholds, and maintenance actions are triggered based on simple conditional logic rather than complex real-time optimization algorithms, reducing computational resource consumption while maintaining operational efficiency.
Solution Approach 2:
The patent uses lightweight, computationally inexpensive threshold-based policies instead of heavy real-time optimization algorithms. These simple decision rules can be executed with minimal computational resources, enabling real-time scheduling decisions without consuming excessive processing power or energy, effectively using 'cheap' computational approaches for time-critical decisions.
Data Source
AI summary
A condition of an asset including one or more subsystems connected in series is monitored. Each subsystem includes one or more components connected in parallel. The asset has one or more jobs. A probability of the asset surviving a predetermined amount of time is determined based on the monitoring and one or more shared resources. The one or more shared resources are configured to be shared between the subsystems. A model is established using a threshold based heuristic maintenance policy. The model is configured to maximize a number of successful jobs that the asset is able to complete based on the determined probability. The one or more shared resources are allocated to the one or more subsystems based on the model.