Toyo says ‘MI' technology will help in HP tire design

HYOGO, Japan — Toyo Tire Corp. has developed artificial intelligence (AI)-based technologies that it claims will allow it to predict the properties of and optimize the material structure of rubber materials more accurately than previously possible and in less time.

Toyo said having this "materials informatics" (MI) technology at its disposal will allow it to "highly optimize" development of high-performance tires while concurrently reducing the time and cost of development.

Toyo said this latest development is part of its proprietary "Nano Balance Technology" rubber-material development platform that the Japanese tire maker has been researching for nearly a decade.

Toyo has been using MI technology based on its accumulated data assets since 2018 "to validate the predictive technology of recipes and physical properties," and looking forward plans to "develop an environment" where it can put its in-house data "to full and effective use."

Toyo noted in material published on the development that since rubber is a composite formed by adding filler and other ingredients to the base material polymer, each ingredient affects the characteristics of the elastomer, thus requiring advanced control in terms of the type and amount, as well as adjustments in the processing method.

Toyo claims using MI technology will allow it to make highly accurate predictions of properties and recipes, a process that will yield efficient material development with minimal testing.

The system implements a "nonlinear regression model," which Toyo said can make expanded predictions that go beyond existing knowledge by importing external information to a database.

The development of MI technology builds on other AI-assisted technologies Toyo has been working on, including a real-time simulation process and "snow prediction" technology under the firm's "T-Mode" tire development platform.

These new technologies integrates computer-aided engineering (CAE) and AI to deliver real-time simulation based on machine learning, Toyo said.

In the conventional development process, design specs are entered and a simulation is run. With the new approach, design specs are revised and the simulation process is performed repeatedly until the calculated performance values achieve the targeted performance, Toyo said.

The new process integrates various data accrued from simulations run thus far in T-Mode, helped by a centralized data management system.

In addition, for the establishment of this technology, Toyo Tire conducted a joint research with Osaka University, in which a hierarchical diagram was developed for transecting tire design specs and tire performance in accordance with design engineering concepts.

In terms of snow traction, Toyo claims newly evolved simulation platform technology makes it possible to predict with a high level of precision the snow-traction performance.

Previously, Toyo Tire established a method for predicting the snow traction performance of a tire in specific types of snow quality (fresh snow, slush) by looking at the shear characteristics of the interface between tire rubber and natural snow, taking into account load and speed factors during driving and braking.

Owing to this, Toyo and its research partners were able to visualize the deformation of individual tread blocks during driving and braking in various environments in which tires are used. This enables the optimum design of tread patterns that suit the environment in which the tires are used, Toyo said.