Sumitomo Rubber taps AI to cut tire analysis time by 90%

Kobe, Japan – Sumitomo Rubber Industries (SRI), a producer of the Dunlop and Falken tire brands, and Japanese AI and IT specialist Fujitsu have developed an artificial intelligence-based technology that is claimed to reduce tire structural analysis time by around 90% while maintaining a high level of accuracy.

In a 3 June statement, SRI said the “AI surrogate model” was developed as part of its long-term digital transformation strategy for tire design and has been validated through a proof-of-concept project.

The technology was applied to the structural analysis of tire deformation under road contact conditions and reduced processing time from about 45 minutes to around five minutes, while analysing approximately 600,000 elements (meshes).

According to SRI, the technology predicts tire performance “with high accuracy and in a short time” by using accumulated finite element method (FEM) analysis results as training data.

The AI model, based on a ‘graph neural network’ (GNN) algorithm, was developed using SRI's tire design expertise and actual design data combined with Fujitsu's AI capabilities.

The companies said the proof-of-concept focused on evaluating deformation behaviour and contact characteristics, including contact shape and pressure distribution between the tire and road surface.

The technology predicted tire-to-road contact shape with an average accuracy of 87.7% compared with conventional FEM analysis, according to the partners.

SRI said the development could accelerate data-driven tire development by reducing the number of design iterations required to determine tire structure and material specifications.

Based on the results, the companies plan to develop a tire design support tool for practical implementation at SRI by April 2027.

The partners also plan to begin testing the technology on a prototype version of Fujitsu's next-generation ‘Arm-based’ Fujitsu-Monaka processor by December 2026, with the aim of optimising inference speed, accuracy and power efficiency.