Reduce one hour of paperwork to one minute, reducing search time by 98%. A long-established manufacturer uses generative AI to create time to engage with "technology and creativity."
Sansha Electric Manufacturing Co., Ltd.
Issues before implementation
- It took 1-2 hours to create the foundation for administrative tasks such as creating meeting minutes and announcements, which took away from the time that should have been spent on planning and specialized review.
- Because factory maintenance records spanning 30 years had been kept on paper, it took a long time to identify past cases from the vast amount of files.
- Manual interpretation and translation work in departments with contact with overseas companies was outsourced or done manually, resulting in cost and communication time lags.
Effects after implementation
- AI-based draft generation reduces writing time from 1-2 hours to just a few minutes, reducing the man-hours required from creation to revision by up to 1/60th
- A field employee with no programming experience created an Excel search system with the support of AI, reducing a one-hour search task to one to two minutes
- Utilizing AI Works enables highly accurate translation, reducing outsourcing costs and improving response speed for departments with overseas contacts.
table of contents
Company Introduction
Sansha Electric Works, Ltd. has been a power electronics manufacturer focused on the technological field of "power conversion and control" since its founding in 1933. With power supply equipment and power semiconductors as its core businesses, the company provides products for a wide range of applications that support society, including the industrial, infrastructure, and energy sectors.
Among these, our "power supplies for surface treatment," which hold the top domestic market share, are essential devices that provide precise electrical control in the "plating process," which increases the wear resistance and oxidation resistance of metal surfaces.
Building on the technological capabilities we have cultivated over many years and the creativity of each employee, we aim to realize our purpose of "advancing society through power electronics and creativity."
Issues before introducing generative AI
For Sansha Electric Works, a company with a history of over 90 years, spreading generative AI throughout the company was a major challenge. Before the introduction, there was no specific know-how, and it was not easy to see AI as something that applied to everyone, especially in administrative work in indirect departments and specialized work at factories.
In particular, at maintenance sites, the vast amount of repair records spanning 20 to 30 years were all managed in paper files. Even when trying to refer to past knowledge, it took a long time to find the desired information due to the analog management, and the utilization of information was hindered to the extent that people sometimes gave up due to the effort required to search.
Effects after generative AI introduction training
After undergoing training, the initial 24 people became promoters, and six months later, the number of participants had increased to 60, and now to over 160 people, which is more than 10% of the total number of consolidated employees.
One notable achievement came from on-site initiative: a maintenance technician with no programming experience partnered with AI to create an Excel-based search program. This enabled information searches that previously took nearly an hour to complete in just one or two minutes.
There have also been major changes in the administrative and sales departments. By using AI to create the base for minutes and announcements, the time it takes to create them has been reduced to just a few minutes. The accuracy of translations has also improved in departments that have contact with overseas companies, and speedy responses have become established through in-house production. The system of "AI as the base, humans to check" has improved the quality of work itself.
Future outlook
Going forward, we aim to go beyond the simple stages of "research" and "writing," and to "embed" generative AI deeply into actual business processes. We plan to link tools such as AI Works with individual business processes and automation services, and move to more in-depth utilization.
Additionally, the Information Systems Department will act as a "hub" to strengthen information dissemination and educational support to other departments. By continuing to listen to the challenges each department faces and proposing AI-based solutions, we will create an environment where all employees can focus on the work they are meant to do and the things they want to do. By combining the technological capabilities of a long-established manufacturer with the speed of AI, we will continue to pursue further improvements in productivity and contributions to society.
