Canon Inc. — Case Study

■Client Information

Company Name	Canon Inc.
Location	3-30-2 Shimomaruko, Ota-ku, Tokyo
Business Description	Development, manufacturing, and sales of printers/office equipment, commercial printing machines, medical systems, cameras/lenses/imaging equipment, network cameras, industrial equipment, materials/coatings, and components

We spoke with members of the Peripheral Products Operations, Chemical Development Center.

Datachemical LAB is being utilized at the company's Peripheral Products Operations, Chemical Development Center. Pictured are Mr. Sakurai (back left), Mr. Watabe (back right), Ms. Wada (front left), Mr. Masuda (front center), and Mr. Muranaka (front right), who are leading the adoption of Datachemical LAB within the organization.

Q1. What themes are you using Datachemical LAB for?

Our division is engaged in the development of materials used in office printers — including toners, photosensitive drums, and rubber rollers — and we are also pursuing new business creation by leveraging these material technologies. In recent years, printers have been required to deliver higher reliability and environmental performance than ever before. We recognized that a fundamental transformation of our material development process was necessary to build an ambidextrous development structure capable of achieving these requirements while simultaneously pursuing new business creation. As part of this initiative, we introduced Datachemical LAB in 2024.

After one year of operation, Datachemical LAB has enabled us to achieve improvements in product quality at an early stage in toner development, strengthen collaboration with the production division, and realize a strong cost-performance ratio. In particular, the ability to simultaneously analyze more than 20 regression methods, combined with the original feature importance analysis method CVPFI, has not only significantly improved development efficiency but also enhanced the reliability of our understanding of the underlying mechanisms. We believe it was precisely because we were able to establish mechanisms that everyone could accept that smooth collaboration with the production division became possible. In the area of new business creation, Bayesian Optimization allows us to leverage our accumulated data to explore materials that meet customer needs across a broad solution space, efficiently.

Q2. What were your reasons for selecting Datachemical LAB?

Since around 2019, a group of enthusiasts interested in MI had been working to generate MI use cases using Python. Through this experience, while we felt strong potential in MI for improving development efficiency, we were also strongly aware that standardizing MI across the organization would require a no-code tool that anyone could use to engage with MI. After evaluating several tools, we selected Datachemical LAB for the following two reasons.

1. A user-friendly interface accessible to everyone

After uploading a CSV containing the dataset, analysis can be performed simply by entering parameters in the order they appear on screen. The tutorials are comprehensive, and we felt that meaningful MI results could be achieved even without expertise in programming or data science — which would accelerate the standardization of MI across the organization.

2. A rich variety of analytical methods

The availability of more than 20 regression methods — both linear and non-linear — with the ability to run them in parallel was a significant advantage. The platform also handles regression and analysis well even with small experimental datasets, which proved to be a good fit for our material development environment.

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

Two key benefits stand out. The first is improved quality and efficiency in material development. In toner development, clarifying the underlying mechanisms enabled us to improve product quality. The ability to take proactive measures also resulted in a dramatic reduction in development lead time compared to conventional methods, allowing us to deliver high-quality products on a stable basis.

The second is the acceleration of MI awareness across the organization. By communicating internally the business impact demonstrated through this and several other MI use cases, we have received growing interest from colleagues expressing a desire to try MI themselves. We believe the outstanding analytical capability and UI of Datachemical LAB are what made it possible to create a flagship use case and build an environment that anyone in the organization can build upon. As a result, we feel the groundwork is now in place to promote MI adoption across the organization as a whole.

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

We are highly satisfied with both the updates and the support since implementing Datachemical LAB. Regular updates ensure that new features and improvements are reflected quickly, allowing us to always access the latest analytical technologies.

The support team is also responsive and thorough, providing prompt assistance with technical questions and issues. For example, they proactively schedule regular sessions to gather our questions and improvement requests. We are grateful that some of our improvement requests have been incorporated into the service in as little as a month.

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

Going forward, we plan to expand the use of Datachemical LAB across even more projects and elevate data analysis skills across the entire organization. In particular, we believe it will prove to be a highly powerful tool for new material development and manufacturing process optimization, enabling more accurate prediction and more efficient development.

The availability of a well-structured educational package also means that even beginners in data science can get up to speed quickly, allowing us to broaden adoption across the organization. This, in turn, will raise data analysis skills organization-wide and establish the foundation for fostering a data-driven culture.

Ultimately, our goal is to advance toward the practical application of automated experimentation and digital laboratory initiatives, dramatically improving both the speed and accuracy of research and development. We expect this to enable us to bring competitive products to market more rapidly.

← Back to Case Studies