Contribution: Introduction to the Carbon Credit Evaluation System Using Generative AI by Osaka Gas Co., Ltd. (Second half)

This article is a contribution from Mr. Eiji Natsuaki, Executive Officer and Head of the Business Creation Division of Osaka Gas Co., Ltd., regarding our company’s efforts to use generative AI for a carbon credit evaluation system. It is the second part of a two-part series.

This is the second part of the article. The first part can be found here:
“Introduction to the Carbon Credit Evaluation System Using Generative AI by Osaka Gas Co., Ltd. (First half)”

1. Approach to the Carbon Credit Domain

Our company aims to achieve carbon neutrality by 2050 and plans to contribute to a reduction of 10 million tons of carbon dioxide emissions by 2030. To achieve this goal, our company sees carbon credits as an effective means. Currently, we are providing carbon credits bundled with gas and electricity, delivering carbon-neutral energy to customers. Going forward, our company is also considering the standalone sale of carbon credits, expanding opportunities for more people to contribute to carbon neutrality.

Figure 1: Image of our carbon credit offerings

Interviews with customers revealed that the three key factors they look for in carbon credits are quality, variety, and price. Therefore, our company is considering handling high-quality, diverse, and affordable carbon credits (Figure 2).

Figure 2: Example of a carbon credit portfolio

The quality of carbon credits is crucial, as low-quality ones could harm the customer’s reputation. Therefore, we will continuously improve the generative AI-based quality evaluation system developed this time to provide customers with carbon credits they can purchase with confidence.

2. Screening of Carbon Credit Projects Using Generative AI

Carbon credit projects are diverse, and their quality varies. Therefore, careful selection is necessary when considering investment. We conduct a detailed risk assessment by comparing with similar existing projects. This method allows us to selectively handle only high-quality credits.

Figure 3: Example of carbon credit project screening

As shown in Figure 3, by evaluating on the same axes as existing similar projects and visualizing the relative quality and risks, we believe it can lead to appropriate and unbiased screening.

3. Future Development Direction

The focus of future development is on increasing diversity and improving accuracy. Carbon credit projects include a variety of types, such as nature-based, emission reduction, and CO2 removal. As new projects and rules are expected to emerge, we aim to develop a system that can flexibly adapt to them. We will also continue to work on improving the accuracy of the generative AI. While perfection is difficult, the goal is to get as close to 100% as possible. In this way, we will continue to develop a more accurate evaluation system while adapting to the ever-changing environment.

4. System Development Using AWS

We have maximized the use of AWS services in developing this system. The scalable infrastructure and advanced generative AI services of AWS enable agile and flexible development. The AWS infrastructure services allow efficient system development and operation by flexibly adjusting resources according to the project scale. Additionally, AWS security features provide “assurance” for the carbon credit evaluation process, which handles highly confidential data. Furthermore, by leveraging the wealth of AWS services, the training and deployment of generative AI models can be done quickly, enabling development that responds to market needs.

5. Conclusion

The carbon credit quality evaluation system developed by us is believed to play an important role in climate change mitigation. However, we do not intend to monopolize this technology, but rather to share it widely and utilize it in collaboration with many others.

Climate change is a global issue that requires a rapid response. Therefore, we prioritize cooperation over competition, aiming to contribute to the sound development of the market as a whole through this technology.

Improving the reliability and transparency of the carbon credit market is the key to maximizing its impact. We plan to further enhance the accuracy and reliability of this system through collaboration with others. By incorporating feedback from various industries and regions, a more fair and inclusive system can be built, contributing to the overall market benefit and accelerating climate change mitigation.

Ultimately, we hope to make this system a common platform for climate change countermeasures. To this end, we will emphasize cooperation with others and aim for the healthy development of the carbon credit market. Time is of the essence in addressing climate change, so cooperation beyond a single company, across the industry and globally, is essential. We will contribute to the realization of a sustainable future through this initiative.

Author

This article was translated by AWS Professional Services Riho Matsui, and Solutions Architect Satoshi Aoyama.