Machine Learning Corrosion Prediction for Metal Containers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for evaluating corrosion in metal containers are time-consuming and require lengthy storage tests, making it difficult to promptly release new products that meet market demands, as they only detect defects after prolonged storage and require extensive man-hours for sample preparation.
Innovation Solution
A machine learning-based deterioration estimation system that uses training data from actual products to predict corrosion in metal containers, including container, contents, and environmental data, allowing for rapid evaluation of deterioration without the need for lengthy storage tests.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If storage tests are conducted to measure corrosion of metal containers, then reliability of corrosion detection is improved, but loss of time increases significantly
Solution Approach 1:
The patent applies preliminary action by measuring electrical resistance of the inner surface coating before storage tests to predict future corrosion levels. This allows the system to anticipate deterioration without waiting for actual storage periods to elapse, thereby resolving the contradiction between reliable corrosion detection and time consumption.
Solution Approach 2:
The patent replaces the mechanical/time-based storage test system with an electrical measurement system. Instead of physically storing products for months and then measuring corrosion, the system uses electrical resistance measurements and machine learning models to predict corrosion, substituting a fast electrical detection method for a slow physical aging process.
2Reliability
If storage tests are conducted for new products, then product safety is ensured, but productivity decreases due to extended development cycles
Solution Approach 1:
The system performs preliminary corrosion assessment by measuring electrical resistance before products enter storage tests. The machine learning model predicts future corrosion based on initial measurements, allowing safety evaluation to be done in advance rather than waiting for storage tests to complete, thus accelerating product development while maintaining safety standards.
Solution Approach 2:
The patent creates a virtual model of corrosion progression using machine learning algorithms that replicate the effects of long-term storage based on short-term electrical resistance measurements. This computational copy of the storage test process allows safety prediction without requiring actual physical storage, thereby improving productivity while ensuring product safety.
3Ease of operation
If electrical resistance measurement is performed to detect coating defects, then ease of operation is improved, but measurement precision is insufficient for predicting long-term corrosion
Solution Approach 1:
The system uses feedback by continuously monitoring electrical resistance measurements and feeding this data into machine learning models that have been trained on actual storage test results. The model adjusts its predictions based on the relationship between initial resistance values and observed corrosion after storage, thereby improving measurement precision for long-term corrosion prediction while maintaining the simplicity of electrical resistance measurement.
Solution Approach 2:
The patent transforms the electrical resistance measurement from a simple defect detection parameter into a predictive parameter for long-term corrosion by changing how the data is processed and interpreted. Instead of using raw resistance values alone, the system incorporates these measurements into machine learning models that predict future corrosion states, thereby enhancing measurement precision without complicating the measurement process itself.
Data Source
AI summary
To accurately estimate deterioration of a product in which contents is held in a metal container. The deterioration estimate system comprises: a predictive model 3 trained mechanically by a training data 1 collected through storage tests of the product; an inputter 5 transmitting data to estimate deterioration; and a transmitter 4 that transmitting a degree of the deterioration of the product computed by the predictive model 3 based on the data transmitted from the inputter 5. The training data 1 includes: data relating to the metal container of the product; data relating to the contents contained in the metal container; data relating to an environment where the product has been stored; and data relating to the degree of the deterioration of the product. The data transmitted from the inputter 5 includes: data relating to the container of a target product to estimate the degree of the deterioration, data relating to the contents of the target product; and data relating to an environment where the target product will be stored.