Skip to content
Home » Biomass Power Technologies Co., Ltd. Innovation in the operation of woody biomass power plants! Announcement of introduction of driving safety support system using large-scale language model (LLM)

Biomass Power Technologies Co., Ltd. Innovation in the operation of woody biomass power plants! Announcement of introduction of driving safety support system using large-scale language model (LLM)

[Biomass Power Technologies Co., Ltd.]
Innovation in the operation of woody biomass power plants!
Announcement of introduction of driving safety support system using large-scale language model (LLM)
*Biomass Power Technologies Co., Ltd.*
Press release: July 5, 2024
**
Innovation in the operation of woody biomass power plants!
Announcement of introduction of driving safety support system using large-scale language model (LLM)
*With the introduction, search time for trouble information has been reduced by up to 125 minutes! An operation safety support system that improves the operational efficiency of woody biomass power plants is now in service! *
* Biomass Power Technologies Co., Ltd. (hereinafter referred to as “BPT”) has recently announced that the Ministry of Economy, Trade and Industry’s
BM Ecomo Co., Ltd. (hereinafter referred to as “BM
It’s called “Ecomo.” ) We are pleased to announce that we have built and launched a security support function called BM-GPT, which utilizes the data accumulated in the biomass power plant operation support system called BMecomo, co-produced with ), using a large-scale language model (LLM). I will do it. *
* 1. Background to the introduction and overview of the system * Compared to thermal power generation that uses fossil resources as fuel, wood biomass power generation has different fuel properties (moisture content and shape).
Because of the different
There were several cases of sudden failures that caused power generation to stop. In order to maintain stable operation, a major challenge was not only to prevent these sudden failures, but also to minimize the impact (shorten the power generation stop time) even if they did occur. .
To overcome this issue, in addition to preventing failures by improving the skills and knowledge of operators, it is also possible to recover from failures in a short time by quickly searching for relevant information when a sudden failure occurs. This is necessary. By combining the data accumulated in “BMecomo” and past
troubleshooting with large-scale language models (LLM), this function has the following system configuration and implementation process to help overcome the above challenges.

* 2. Overview of security support function “BM-GPT” *
The security support function “BM-GPT” incorporating LLM consists of the following functions.
I. Search function for each process value
A function to search process value data stored in the existing BMecomo system database using LLM
II.Maintenance record search function
A function to search maintenance records in PDF format stored on internal servers using LLM
III. User interface (UI) development
Chatbot development as a UI to perform the above search

* 3. Effects of introducing the security support function “BM-GPT” * The introduction of this function will help to quickly recover from problems at our wood biomass power plants and minimize power generation outage time. Specifically, the time taken to search for information when a problem occurs is up to 125% before and after implementation.
The effect of reducing the number of minutes/case has been proven. In addition, by using this function, information can be shared smoothly even when it comes to individualized tasks and responses that are difficult to reduce, and by improving the skills of operators and leveling their skills, 72.
It is expected that the work time will be reduced by an equivalent amount of time.
In the future, by combining this function with various predictive models that are currently being developed, it will be possible to realize even more support functions such as notifications of equipment inspection timing and operational advice using chatbots, which will lead to data-driven technology. The qualitative effects of power plant management are expected to be further improved operational awareness and innovation in on-site operations.
* Four. Overview of Biomass Power Technologies Co., Ltd.*
* *About the wood biomass power plant operation support system “BMecomo”* “BMecomo” is a biomass power generation system that utilizes the latest technologies such as big data analysis, AI, and IoT.
This is an operation support system aimed at supporting operational management, equipment management, and business management at a facility. latest technology
We aim to maximize the profits of biomass power plants by utilizing G and thorough “visualization”.
“BMecomo” service introduction page: https://bmecomo.com/

-contact information-
Biomass Power Technologies Co., Ltd.
Business Administration Department Public Relations
TEL:0598‒67‒2561
MAIL:bpt‒shared@bpt.co.jp
*About details about this release*
https://prtimes.jp/main/html/rd/p/000000008.000114663.html

*Download press release materials*
https://prtimes.jp/im/action.php?run=html&page=releaseimage&company_id=114663&release_id=8




Copyright © PR TIMES Corporation All Rights Reserved.