Machine Learning Prediction of Tire Tread Compound Dynamics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The tire manufacturing sector faces challenges in determining the dynamic properties of rubber compounds for tire treads, such as tan δ and E′, which are crucial for energy dissipation and rolling resistance, due to lengthy and costly laboratory testing processes, leading to increased lead times and variability in data.
Innovation Solution
A predictive method using machine learning, specifically an artificial neural network, is employed to simulate laboratory tests, normalizing and preprocessing data to accurately estimate dynamic properties without physical testing, reducing variability and improving predictive precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive laboratory testing is performed to validate composite formulations and determine dynamic properties, then measurement precision and reliability are improved, but loss of time and productivity deteriorate
Solution Approach 1:
The patent creates a virtual copy of the laboratory testing process through machine learning models that replicate the behavior of rubber compounds. Instead of physically testing each formulation, the system uses trained ML models to predict dynamic properties (tan δ, E′) based on compositional data, effectively copying the outcome of extensive laboratory testing without the time cost.
Solution Approach 2:
The patent performs preliminary action by training machine learning models on historical laboratory data before actual product development. This pre-trained knowledge base allows rapid prediction of dynamic properties for new formulations without requiring extensive new laboratory testing, thus reducing validation time while maintaining precision.
2Manufacturing precision
If iterative experimental campaigns are conducted to optimize formulation, then manufacturing precision is improved, but loss of time and productivity worsen
Solution Approach 1:
The patent replaces the mechanical system of physical experimentation with an information-based system. Machine learning algorithms process compositional data and predict performance outcomes, substituting the iterative physical testing process with computational analysis that achieves the same optimization goal much faster.
Solution Approach 2:
The patent changes the approach from varying physical formulations through repeated testing to changing computational parameters in the machine learning model. By adjusting input features and model parameters based on historical data patterns, the system achieves formulation optimization without the time cost of iterative physical experiments.
3Reliability
If multiple validation steps are performed in the laboratory, then reliability of dynamic properties determination is improved, but device complexity and operational difficulty worsen
Solution Approach 1:
The patent merges multiple separate validation steps into a single integrated machine learning prediction process. Instead of performing sequential laboratory tests for different dynamic properties, the system uses a unified ML model that predicts all required properties (tan δ, E′) simultaneously from compositional data, reducing process complexity while maintaining reliability.
4Measurement precision
If extensive laboratory testing is conducted to evaluate performance, then measurement precision is improved, but loss of substance and cost worsen
Solution Approach 1:
The patent uses virtual copying through machine learning models to predict performance outcomes without physically consuming raw materials. The ML system analyzes compositional data and predicts dynamic properties, eliminating the need to manufacture and test physical samples for each evaluation, thus preventing material loss while maintaining measurement precision.
Data Source
AI summary
The present invention refers to a computer implemented predictive method based upon machine learning for the development of composites for tyre tread compounds. The method comprises the following steps: providing a raw data database, namely, a dataset consisting of recipes for already existing composites and of corresponding known dynamic properties, to be used as a reference; normalizing the data contained in the raw data database according to an iterative procedure; pre-processing the normalized data by means of Data Mining in order to eliminate aberrant data and to add new fictitious ingredients relating to specific categories of actual ingredients; training an algorithm based upon automatic learning by means of the pre-processed data; applying said trained algorithm to a set of experimental data that are representative of the recipe of the composite to be tested, for the prediction of the dynamic properties of said composite to be tested.