Summary
Nippon Shokubai is a Japanese chemical company with the largest share of the global market for polymeric polymers. In December 2022, Nippon Shokubai implemented Cognite Data Fusion®, the leading Industrial DataOps platform, to increase productivity at plants that produce chemical materials. The system was first rolled out at Himeji Works (a plant for acrylic acid and superabsorbent polymers) to improve data discovery workflows for on-site workers searching for equipment-related data for operations, maintenance, and other tasks.
With Cognite Data Fusion®, the company significantly reduces the amount of time spent on this task and improves productivity through predictive maintenance and other advanced analysis. The company plans to expand the system to overseas plants in the U.S., Europe, and other regions in the next deployment phase.
The Vision: More than just a data lake
Digital Transformation (DX) is sweeping across the heavy-asset industry. Specifically, organizations are moving beyond simply digitizing data to address contextualization, the idea of organically connecting various digitized data, analyzing it, and effectively utilizing the insights gained.
Implementing a contextual data foundation has an exceptionally significant impact. It improves operational efficiency, enables the execution of more sophisticated business models, and holds the potential to reform work styles. As such, it is no surprise that many companies in the manufacturing sector, including Nippon Shokubai, are actively embracing contextualization.
In 2021, the company announced a long-term vision titled "TechnoAmenity for the future," outlining its aspirations for three transformations by 2030: "Business Transformation," "Environmental Transformation," and "Organizational Transformation." Additionally, in 2022, to accelerate the realization of these transformations, they introduced the "DX Vision," pushing for the swift establishment of an environment where all employees can leverage digital technology and data.
Challenge
Data contextualization requires significant effort
Realizing the DX vision involves considerable effort in data organization and processing, Nippon Shokuba appointed Mr. Hiroki Nakagawa as the head of the DX Promotion Headquarters in April 2022. Before this role, Mr. Nakagawa served as the Production Management Center Chief at the Himeji Manufacturing Plant, driving operational reforms in the production field as part of the plant's activities.
As part of his work at the Himeji Manufacturing Plant, Mr. Nakagawa led the "Big Data Utilization Working Group.” It was within this group that Mr. Nakagawa first had the idea that "analyzing and processing operational data output by the Integrated Production Control System (DCS) and various event data might quickly uncover the seeds of hidden issues, enabling 'predictive maintenance."
Mr. Nakagawa searched for an optimal tool to bring these new ideas to fruition. He found several, some of which progressed to trial implementations. Still, all failed for the same reason—the immense effort required for data organization and processing before deploying the tools proved too burdensome for the production floor.
Adding to the challenge, the production floor at the Himeji Manufacturing Plant had many "data silos." While the operational data from DCS, maintenance history/fault information data, inspection photos, etc., were digitized, each type of data was stored in systems optimized for specific use, making seamless handling on a single screen impossible. Piping and Instrumentation Diagrams (P&ID) remained in PDF format, and specifications for equipment, among other documents, were still stored in files.
Bringing all of this data together was hard enough. Adding context to make this data meaningful to on-site workers proved very time-consuming.
Solution
A digital twin to oversee factories across Japan
The Himeji Manufacturing Plant sought to address its data silos with an “information integration platform,” but finding a solution proved elusive. Seridepidiusly, however, Mr. Nakagawa received an email about "predictive maintenance" and "data utilization" and attended a Cognite webinar showcasing how Asahi Kasei, a comprehensive chemical manufacturer involved in similar businesses as Nippon Shokubai, constructed a digital twin of its operations using Cognite's Industrial DataOps platform, Cognite Data Fusion®. When Mr. Nakagawa saw how easy it was to click on equipment, instrument symbols, or piping in an interactive P&ID and find all associated data in one place, he knew this was precisely what he had been looking for.
Following this, Mr. Nakagawa did a deep dive into Cognite Data Fusion® and competing tools and even explored proposals for custom tools from consulting companies. According to him, "Cognite Data Fusion® clearly surpassed others. The speed at which we could access the desired information was overwhelmingly fast." With this conviction, he secured the agreement of key stakeholders, including executives at the Himeji Manufacturing Plant and headquarters, conducted a Proof of Concept (PoC) to verify functionality and performance, and officially adopted Cognite Data Fusion® in December 2022.
Key Results
The initial implementation, supported by Cognite's dispatched engineers, took three months. They selected 12 out of approximately 70 production plants at the Himeji Manufacturing Plant and imported and contextualized key operations data (equipment, operational, drawing, maintenance, etc.) into Cognite Data Fusion®. According to Mr. Nakagawa, "The feedback from the field was very positive, and numerous requests were received for the incorporation of additional data and the expansion to all plants."
In response, the Himeji Manufacturing Plant has appointed a person to be in charge of Cognite Data Fusion® implementation and is currently progressing with the expansion to all 104 facilities (including plants and utility facilities) without the assistance of Cognite engineers (Figure 1). Mr. Nakagawa said, "As of February 2024, the progress is about 85%, and we plan to complete everything by the end of March 2024."
The main task currently underway is contextualizing information imported into Cognite Data Fusion®, with approximately 80% of this work supported by the Cognitie AI features implemented in Cognite Data Fusion®. While effort is still required for the remaining tasks, the overall time commitment is significantly less than for other tools.
The company estimates implementing Cognite Data Fusion® will reduce the time spent searching for plant-related data in the production field by approximately 9,000 hours per year (Figure 2). As a result, valuable human hours in the production field are saved. "In the future, Japan will see a decrease in the working population. It is necessary to establish a system to cope with this," notes Mr. Nakagawa.
The company plans to strengthen its relationship with Cognite and expand the scope of its use of Cognite Data Fusion® in three directions.
- Firstly, create a digital twin akin to Google's "Street View" for factories by importing point cloud data and panorama photos into Cognite Data Fusion®.
- Secondly, utilize robots to monitor equipment and collect data to reduce the workload for on-site engineers.
- Thirdly, deploy Cognite Data Fusion® in overseas factories so that equipment, operational, drawing, and maintenance data can be accessed from Japan, enabling faster issue response.
