The development of IZANA started in 2015. Colleagues were working on a way to promptly identify failures in already delivered power generation equipment when they noticed the possibility of applying a machine learning model developed by our Technology Research Center.
The name IZANA is derived from the Japanese verb "izanau," meaning "to invite," reflecting a desire to invite (izana-u) facilities, businesses, and social contribution to move in a positive direction.

When power generation facilities unexpectedly stop working, this not only causes economic damage but also disrupts electricity supplies. Customers asked us to develop systems that add to the reliability we have built as a manufacturer.

As the development of IZANA progressed, the team started to feel confident that this could become a marketable product. However, as generic technologies are normally only good for solving superficial issues, they needed to look deeper. "We went back to the shop floor over and over again to seek out the real problems and incorporated customers' feedback into the functionality," reminisces a senior engineer in the Corporate ICT Group.

In the conventional problem solving process, staff are dispatched to the site in question where they check data and propose a plan for improvement. However, that takes a lot of time and effort.
Therefore, using the SHI Group's ICT/IoT shared platform (SHICuTe*), the team built a system that can constantly collect operational data and share it in real time while ensuring security. This enables investigators to rapidly understand the situation and propose solutions.

Trial runs of IZANA were conducted from 2018. Over a period of 3 years, a more reliable algorithm was developed, and the system was refined by carefully picking up the needs from the shop floor.

Nonetheless, that did not mean it was immediately accepted by everyone. Proud of their skilled operators, some customers were dismissive, "Our people are experts, we don't need it."
In contrast, businesses worried about technology transfer welcomed the introduction of the new system, which was given fresh impetus by the dawning of the AI era.

One senior engineer in the Energy & Environment Group describes the current status as follows. "IZANA is functional to some extent but is not yet finished. In future, we are aiming for a system that AI can run automatically even when no operators are present."

Our industry features very few operating support systems developed by the facility makers themselves. Most are general systems managed by IT companies.
In such a situation, it was the strong desire of the IZANA development team to "get close to the shop floor and make something really useful" that was key in driving the project to success.

Former boiler designers, factory workers and others from various backgrounds put their shop floor experience to good use and contributed to the development.
"With IZANA, tasks should become more worker-friendly, leading to a more considerate working environment," believes a senior research engineer.

IZANA was finally commercialized in summer 2023. There were many times that the project seemed to be on the verge of collapse, but colleagues stepped up each time to take the work forward.
For that reason, the project simply had to be taken through to its conclusion. For all the colleagues who offered their help, and for society that needs this technology...

*IZANA and SHICuTe are registered trademarks of Sumitomo Heavy Industries, Ltd.