Fujitsu accelerates AI integration in manufacturing digital solutions

September 2, 2025

By: Masanobu Nishizuka, Staff Editor, SEISANZAI Japan

Fujitsu is expanding the use of artificial intelligence (AI) within its manufacturing digital solutions as part of its Uvance business, which aims to solve social issues through cross-industry collaboration.

On July 16, the company hosted its annual event, “Fujitsu Uvance Update,” to report on the latest progress. Yoshinami Takahashi, Corporate Executive Officer, Corporate Vice President and Chief Operating Officer (COO), emphasized: “We are entering an era where AI supports all areas of business and works in collaboration with people.”

AI agents for supply chain management

Yoshinami Takahashi, Corporate Executive Officer, Corporate Vice President, COO in charge of Solution Services, and Head of Global Solutions, Fujitsu

One of the highlights was the use of AI agents—autonomous systems that make decisions and take actions toward goals set by humans.

Fujitsu demonstrated how multiple AI agents could be applied to supply chain management (SCM). In the example, several specialized AI agents managed inventory on the SCM platform. When a risk of stock shortage emerged, a human operator instructed the system to “prioritize cost in determining the optimal solution.” Four task-specific AI agents—covering procurement, inventory control, manufacturing, and sales—then presented their respective proposals.

An AI “orchestrator” coordinated these proposals and delivered an overall optimized decision in just a few seconds.

Preventing the “black box” problem

AI agents exchanging opinions on SCM (courtesy of Fujitsu)

The agents’ reasoning processes and their discussions with one another are recorded in natural language, addressing concerns over the “black box” nature of AI decision-making. This also enables companies to understand the tendencies of the agents’ thought processes. By aligning these tendencies with corporate culture, companies can guide and adjust the agents’ decision-making.

A Fujitsu developer explained: “It is like a consultation between a staff member and a supervisor. Ultimately, the decision-making flow among AI agents begins to resemble that of a corporate organization. The more it is used, the more the AI agents learn and adapt to the unique culture of each company.”