Sumitomo Chemical Co., Ltd., at nearly $17 billion in annual revenue, is one of Japan’s largest chemical manufacturers and has roughly a decade of digital transformation experience under its belt. The chemical company has operations that span 66 countries, with approximately 6,500 employees globally.
Sumitomo Chemical runs a diverse portfolio of businesses, including its agriculture and life sciences business, mobility, information and communications technology, advanced medical, essential, and green materials business. About a decade ago, Sumitomo Chemical recognized the need to modernize and streamline the operations for those businesses— including transformations with IoT, research and development, supply chain, cybersecurity, and the move toward becoming an AI-native organization.
That transformation began with IoT and spans to its current “DX NEXT empowered by AI” initiative, which was launched earlier this year. Over the decade, the company successfully positioned its digital transformation as a core element of its strategy, integrating it deeply into its three-year business plan, with the slogan Leap Beyond: Returning to a Growth Trajectory, for fiscal years 2025 to 2027.
“We have always sought ways to improve and to transform the business and implement AI digitally,” said Erwin Eimers, chief information officer and chief information security officer at Sumitomo Chemical of Americas, an American subsidiary of Sumitomo Chemical. In early 2024, Eimers established the Enterprise AI Council, a multi-disciplinary AI governance body for the Americas, as a part of the broader company’s digital transformation initiatives. The
body brings together leaders from IT, cybersecurity, research, manufacturing, HR, and individual business units in the Americas to review major AI initiatives, share learnings, and promote best practices to improve success. The group also ensures these efforts align with Sumitomo Chemical’s broader strategy, and it evaluates the associated risks to privacy, data protection, security, and regulatory compliance.
Currently, according to Eimers, they are focused on implementing AI for three key use cases: improving productivity, supply chain management and forecasting, and implementing AI on the production lines. And this year, the company merged its Digital and Data Science Innovation Department with the IT Innovation Department to form the new “DX Acceleration Office.” The centralization is designed to drive company-wide DX initiatives while providing enhanced senior management support. Sumitomo Chemical’s AI vision encompasses not merely the adoption of AI as supplementary technology but its integration as a fundamental element in product development, manufacturing optimization, and customer engagement strategies.
One of Sumitomo Chemical’s most visible AI implementations is dubbed ChatSCC, an internal generative AI service utilizing ChatGPT technology that was launched in October 2023. This platform represents a comprehensive enterprise AI deployment, making advanced AI capabilities available to all 6,500 employees across the organization.
The development and deployment of ChatSCC underscores Sumitomo Chemical’s stated commitment to practical AI implementation with measurable business outcomes. During preliminary verification processes, for instance, the company tested ChatSCC across approximately 200 typical business operation patterns, achieving efficiency improvements of up to 50% in various scenarios, including document creation, email drafting, summarization, idea generation, and program development.
Advanced Security Operations Center (SOC) Implementation
Sumitomo Chemical’s Agentic AI implementation in cybersecurity focuses on three core objectives: enhanced threat detection capabilities, improved incident response times, and support for lean security teams operating across multiple time zones and geographic regions. The system functions as an intelligent assistant that aggregates data from various sources, provides contextual analysis, and presents actionable recommendations to human analysts, enabling faster and more accurate decision-making processes.
For example, the system has been configured with varying levels of autonomy depending on risk assessment – it can autonomously lock user accounts when suspicious activity is detected, but requires human approval for actions affecting service accounts or critical infrastructure components. “We use it as our single pane of glass for monitoring IT and OT security. It pulls quick summarizations together detailing what any potential issue may be and saves us an enormous amount of time in investigation,” said Eimers.
To achieve this, they selected Stellar Cyber, a platform capable of ingesting data from a wide array of sources—endpoint detection and response systems, firewalls, email security, and more. Unlike traditional rule-based systems, Stellar’s agentic AI could autonomously triage alerts, summarize incidents, and even take specific actions, all while learning from feedback to improve over time.
The deployment was methodical. The team started by running the AI in observability mode, allowing it to analyze and report on incidents without taking direct action. This cautious approach enabled the team to fine-tune the system, ensuring it understood the data and made sound recommendations. Only after extensive training and validation did they allow the AI to act autonomously in specific scenarios, Eimers explained.
Throughout the process, the company learned valuable lessons. First: clean, comprehensive data is essential, as the “garbage in, garbage out” principle remains as relevant today as it was when the term was coined 80 years ago. They also found that starting with passive monitoring helped build trust in the system and allowed for gradual, safe adoption of autonomous actions. The agentic AI quickly proved its worth, making triage faster and more efficient, and freeing up human analysts to focus on more complex issues. “One of the most beneficial aspects of agentic AI in cybersecurity is that it learns, compared to pure rules-based systems, AI remembers changes and doesn’t have to be repeatedly told,” Eimers said.
Materials Science and Chemical Informatics
Sumitomo Chemical has also implemented AI technologies in materials science research through partnerships with companies like Hitachi High-Tech Solutions for chemical informatics applications. These AI systems analyze vast databases of chemical compounds to identify promising candidates for specific material properties, significantly expanding the scope of materials research while reducing development timelines.
The AI here enables researchers to search among millions of potential compounds, increasing the probability of finding functional materials from approximately 1 in 100,000 using traditional methods to 1 in 100 with AI assistance. This capability acceleration supports Sumitomo Chemical’s development of advanced materials for semiconductor applications, energy storage systems, and sustainable chemical products.
Smart Manufacturing and IoT Implementation
Sumitomo Chemical’s smart manufacturing initiatives represent a comprehensive integration of IoT technologies, AI-driven analytics, and automated systems across production facilities. The company’s IoT strategy, initially launched in Singapore in 2016 with support from the Singapore Economic Development Board, has expanded to encompass predictive maintenance, real-time production optimization, and automated quality control systems.
The implementation includes comprehensive sensor networks throughout production facilities, enabling real-time monitoring of equipment performance, environmental conditions, and product quality parameters. AI algorithms analyze this continuous data stream to identify patterns indicative of potential equipment failures, quality issues, or process optimization opportunities, enabling proactive rather than reactive manufacturing management.
“Much of this is still a work in progress,” said Eimers, detailing how data limitations hampered a digital twin deployment, and the need for ever more sensors to be deployed. “It’s not easy to put AI in your production line. You need first to deploy all of the sensors, collect the data, and build a data warehouse before you can get much value out of it,” he said.
Global Supply Chain Digitization, Predictive Maintenance, and Equipment Optimization
Sumitomo Chemical has implemented comprehensive digital technologies throughout its global supply chain operations, creating real-time visibility and optimization capabilities across manufacturing, distribution, and customer fulfillment processes. The system integrates production planning, inventory management, and logistics coordination through AI-driven analytics that adapt to demand fluctuations and supply chain disruptions.
Advanced demand forecasting capabilities utilize machine learning algorithms to analyze market trends, customer behavior patterns, and external factors that influence product demand. This predictive capability enables more accurate production planning and inventory optimization, reducing costs while improving customer service levels.
Advanced predictive maintenance capabilities represent a significant area of technological advancement for Sumitomo Chemical. “We are very aggressive on Industry 4.0. We are not fully there yet, but we are on the road to getting there.
The company has implemented AI-driven systems that analyze equipment data, including vibration patterns, temperature fluctuations, oil analysis results, and operational parameters to predict potential failures before they occur.
Sumitomo Chemical’s digital transformation investments have contributed significantly to the company’s financial recovery and growth trajectory. The company’s strategic focus on
digital transformation has supported improvements across key financial metrics, with return on invested capital reaching 2.2% in fiscal 2024 and targeted to reach 6% by fiscal 2027 under the current business plan. These improvements reflect both operational efficiency gains from digital technologies and revenue growth from digitally enabled new business models.
Today, Sumitomo Chemical of Americas continues to expand its use of agentic AI, exploring new applications in supply chain forecasting and production line optimization. The journey is ongoing, but the company’s experience demonstrates the power of a thoughtful, data-driven
approach to deploying advanced AI technologies. “We have big plans to continue driving these types of efforts forward and improve from the gains already made,” Eimers said.
