Ohara Inc. — Case Study

■Client Information

Company Name	Ohara Inc.
Location	1-15-30 Koyama, Chuo-ku, Sagamihara City, Kanagawa Prefecture
Business Description	Manufacturing and sales of glass materials for optical and electronics equipment

We spoke with K.O. from the Research & Development Center.

Datachemical LAB is being utilized at the company's Research & Development Center. We spoke with K.O., who is leading the internal adoption of Datachemical LAB.

Q1. What themes are you using Datachemical LAB for?

Our R&D division conducts research and development of glass and glass-ceramics that contribute to a wide range of fields, including optics, electronics, environment, and energy. The pace of technological innovation in recent years has been remarkable, and there is a growing demand for the development of next-generation glass materials with an eye to the future.

In pursuit of developing superior glass materials more rapidly, our division leverages Datachemical LAB's Regression Analysis and Bayesian Optimization features for the exploration of glass compositions aimed at imparting new properties or improving performance, as well as for physical property prediction. Composition development is a core technology that our company has cultivated over many years, and we have access to tens of thousands of glass material data records. We believe Datachemical LAB is ideally suited for discovering new possibilities in glass development.

Q2. What were your reasons for selecting Datachemical LAB?

We selected it based on several advantages: a rich variety of analytical methods, a strong track record of adoption by companies and research institutions, ease of data visualization, no-code operation with a simple interface, and excellent cost performance.

In particular, developers in the chemistry and materials field do not necessarily possess strong IT skills, and we feel it would be a high hurdle to expect all development personnel to perform computational processing through programming. Being able to engage with data science easily — thanks to the straightforward operability, along with well-developed tutorials and a comprehensive help site accessible to everyone — is something we greatly appreciate.

Furthermore, the abundance of analytical methods, combined with the ability to run simultaneous analyses and easily compare the results of various modeling approaches, means that users can apply analytical results to their development work with genuine confidence. This was also one of the reasons for our selection.

Q3. What benefits have you seen since introducing Datachemical LAB?

We actually used it on a glass composition development project. That project had an extremely tight development timeline and required us to simultaneously achieve both performance targets and intellectual property requirements. Glass development inherently requires a person to conceive a composition and then physically produce a glass sample with that composition through experimentation. Making even a single glass sample takes time, and as the number of experiments increases, so does the physical burden on personnel.

By utilizing Datachemical LAB's Regression Analysis and Bayesian Optimization, we were able to identify a glass that met our requirements with fewer experiments than conventional methods, which ultimately contributed to shortening the development timeline.

In addition, through ongoing use of Datachemical LAB, knowledge and understanding of Materials Informatics (MI) has been gradually spreading throughout our R&D organization as a whole. Because data visualization and analysis are intuitive and easy to understand visually, it has also proven highly effective as educational material for helping researchers themselves develop an understanding of data science. We are now seeing a growing number of people expressing interest in applying it to their own development projects.

Q4. How do you feel about post-implementation updates and our support? 

New features and improvements are released at a high frequency, and we have the impression that the platform is becoming increasingly convenient and user-friendly. We would be delighted to see the latest analytical features continue to be incorporated going forward.

Regular check-in meetings ensure that we receive thorough and prompt follow-up regarding the latest updated information as well as our questions and requests, and we are very satisfied with this support.

The help pages also go beyond simple feature explanations, providing detailed commentary on data analysis — making them highly useful as educational and explanatory resources as well.

Q5. What are your plans for expanding the use of Datachemical LAB internally?

Since our company has a database of tens of thousands of glass materials, we would like to actively leverage that data and continue using Datachemical LAB for composition development. Moreover, beyond composition development alone, we would like to integrate it with our ongoing manufacturing process digitalization (DX) initiatives to address challenges in a coordinated manner across materials, design, and manufacturing.

To achieve this, we recognize the need to broaden awareness and understanding of data science within the company. As we work to develop data-literate personnel, we would also like to use Datachemical LAB as training material. While advances in technology have made data collection easier, data science is essential for effectively leveraging the large volumes of data collected. Datachemical LAB offers intuitive data visualization along with comprehensive tutorials and help pages, making it well-suited as educational material that even beginners can understand.

← Back to Case Studies