Adaptive Time Scale Data Analysis for Manufacturing Loss Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current productivity analysis systems face challenges in analyzing data fluctuations with fine time scales from manufacturing data, leading to enormous data and calculation amounts, making it impractical to determine production loss factors effectively, and fail to provide actionable insights for improving work processes based on analysis results from combined data with different time scales.
Innovation Solution
A productivity improvement support system that includes a time scale setting unit to switch data scales based on state fluctuations, a loss analysis calculation unit to analyze production loss factors using analysis models, and a recommended work selection unit to suggest improvements based on production loss factors, allowing for efficient data analysis and work instruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data is always acquired and analyzed with a fine time scale to detect data fluctuations, then measurement precision is improved, but the quantity of data and calculation amount become enormous
Solution Approach 1:
The patent applies dynamics by making the time scale adjustable rather than fixed. The time scale setting unit dynamically changes the time scale based on the analysis requirements and data characteristics, allowing the system to use fine time scales only when necessary for detecting data fluctuations while using coarser time scales for normal operation, thus resolving the contradiction between measurement precision and data quantity
Solution Approach 2:
The patent changes the time scale parameter adaptively based on the analysis needs. When data fluctuations need to be detected, the system switches to a finer time scale; when such detection is not required, it uses a coarser time scale. This parameter change approach allows the system to maintain measurement precision when needed while significantly reducing the overall data quantity and calculation burden
2Measurement precision
If data is always acquired and analyzed with a fine time scale to detect data fluctuations, then measurement precision is improved, but the amount of calculation becomes enormous
Solution Approach 1:
The system dynamically adjusts the time scale based on whether data fluctuation detection is required. By making the time scale adaptive rather than constantly fine, the calculation amount is significantly reduced while maintaining the capability to detect fluctuations when needed, thus resolving the contradiction between measurement precision and calculation power
Solution Approach 2:
The time scale parameter is changed adaptively based on analysis requirements. The system uses finer time scales only when data fluctuations need to be detected, and coarser time scales otherwise, thereby maintaining measurement precision when necessary while dramatically reducing the overall computational burden
3Measurement precision
If data is acquired with a fine time scale, then measurement precision is improved, but it is practically impossible to calculate due to enormous data and calculation amounts
Solution Approach 1:
The patent makes the time scale dynamic and adaptive rather than fixed at a fine level. The time scale setting unit adjusts the time scale based on whether fluctuation detection is required, enabling the system to maintain measurement precision when needed while ensuring calculability by using coarser time scales during normal operation, thus resolving the contradiction between measurement precision and calculability
Solution Approach 2:
The system changes the time scale parameter adaptively based on analysis requirements. By using finer time scales only when necessary for detecting data fluctuations and coarser time scales otherwise, the system maintains measurement precision when needed while ensuring that the data and calculation amounts remain manageable and calculable
Data Source
AI summary
Improvement of work is instructed based on an analysis result obtained by combining data having different timescales. A productivity improvement support system includes a time scale setting unit configured to, when 4M data having different time scales acquired from a target device contains data that satisfy a condition for detecting a state fluctuation, switch time scale of the 4M data to time scales according to a state fluctuation, a loss analysis calculation unit configured to analyze a production loss factor by using analysis model data in which the production loss factor of the target device when the condition is satisfied is determined, and a recommended work selection unit configured to select a recommended work when the production loss factor occurs from one or a plurality of recommended works by using recommended work data stored in association with the production loss factor.