New RAG-based system helps engineers quickly access technical knowledge and supporting evidence for faster, data-driven tyre development
The Yokohama Rubber Co., Ltd. has developed a proprietary generative AI system designed to support decision-making throughout the tyre development process. The company began full-scale operation of the new system in August 2026, integrating Retrieval-Augmented Generation (RAG) with its internal technical knowledge base.
The system enables Yokohama Rubber’s development teams to search a large collection of internal technical documents and receive responses based on relevant information contained in those materials. These documents include regulatory information, procedure manuals, technical reports and case studies that are used during product and material development.
The new technology expands the practical application of the company’s HAICoLab framework, which was established in 2020 to promote collaboration between humans and AI. Yokohama Rubber has already applied AI to areas including rubber compound property prediction, tyre performance forecasting, new compound generation and mould design.
Under the new system, engineers can enter questions based on their specific development objectives and circumstances. A generative AI system then identifies relevant technical information and prepares responses using the company’s internal documentation.
An AI agent incorporated into the system can also interpret the intent behind engineers’ questions and repeatedly perform search planning, information retrieval and result evaluation. This process is designed to improve the relevance and accuracy of the information presented to development teams.
A key feature is the system’s ability to provide links to the original technical documents used to generate its responses. Engineers can therefore review the underlying information, assess its suitability and use the verified knowledge in their technical evaluations and development decisions.
Yokohama Rubber said the system addresses the challenge of efficiently using the extensive domain knowledge accumulated across its technical documentation. By combining this knowledge with generative AI, the company aims to make technical information more accessible during the development of new materials and tyres.
The initiative forms part of Yokohama Rubber’s wider digital transformation efforts. Alongside developing AI applications, the company is also training DX personnel capable of applying the HAICoLab framework and related technologies.
The company plans to continue improving the system and using its accumulated data and technical knowledge to support the development of innovative tyres, materials, processes and services.