”Office Yamamoto” Company Overview

In April 2026, the acquisition of Toyo Seat by Tachi-S was announced. This may not just be a story of the integration of domestic seat manufacturers. It could signal the full-scale reorganization of the Mazda supply chain and may be a symbolic event in forecasting Mazda's future in 2030. There is also a possibility that similar integrations will accelerate in areas beyond interior components. This paper attempts to envision the changes in Mazda by 2030 from four perspectives: 1) the transformation of the Toyota-Mazda relationship, 2) the realm of the Mazda brand, 3) the power dynamics of Tier 1 interior suppliers, and 4) the elimination and reorganization of Tier N suppliers.

Automation of Specialized Translation With the emergence of AI translation, it is known that automatic translation has advanced to a level of practicality for everyday conversation, although it may lack the naturalness of a native language. Additionally, these AI translation tools support multiple languages, and the number of languages they can handle has expanded from around 30 to nearly 100. However, in specialized translations where technical terminology is frequently used, such as patent documents, technical materials, and contracts, the ability to accurately translate the meanings of specialized terms from the text is still considered insufficient. Even highly skilled interpreters and translators can encounter cases where they struggle with the issue of "not understanding the meaning of Japanese," particularly concerning specialized terminology. This indicates that even with high-performance AI translation, similar challenges remain, which can be seen as issues with automatic translation. Here are some proposals regarding methods to resolve this.

We will introduce a theme that involves the automation of injection molding mold design using an AI (Cause-Effect Analysis Module). I had the opportunity to meet with the head of a facility that accidentally engages in the design and manufacturing of injection molds. They expressed concerns about the "shortage of mold designers" and wondered if there was a way to automate the process to some extent. It is well known that the design of injection molds heavily relies on product information (CAD data) and that the structure of undercut areas can become quite complex. Additionally, when designing injection molded products, it is essential to conduct the design work while being aware of the structure of the injection mold. Therefore, we have decided to focus on the automation of injection mold design as our theme.