AI in specialty chemicals: Lubrizol's fusion teams, cloud data, faster tests
Updated
Illustrative photo
What Lubrizol is building
Lubrizol is putting together a digital base that links business processes, modern data platforms and cloud infrastructure. The aim is to support its research and development teams. Girish Rao, vice president of information technology at the specialtychemicals manufacturer, describes the plan as giving researchers tools to test ideas faster and make better-informed decisions.
The company wants to pair decades of accumulated technical data with the expertise of chemists and engineers. Those specialists understand customer needs and the physical limits of the work. Data scientists sit alongside them to spot patterns, test new hypotheses and model possibilities before anyone runs a physical experiment.
Rao stresses that the point is not to replace scientific judgement but to give scientists better tools.
Fusion teams start small
The three groups involved approach problems differently. Scientists prioritise experimental validity, data scientists look at model performance, and IT teams focus on platform security and scalability. Lubrizol brings them together in what it calls fusion teams, aligned around shared objectives. The method starts with small pilots rather than a big rollout. Rao says trust between team members is the turning point: once people know what value they are about to create, things go a lot better.
Cloud instead of own data centres
Cloud infrastructure gives Lubrizol flexibility to reach computing resources and modern tools. The company avoids the substantial investment needed to build and maintain its own data centres. Rao points to sustainability as an area where these capabilities could prove especially valuable. Digital modelling lets researchers evaluate more formulations and explore materials such as starch and corn-based alternatives. It also lets them assess trade-offs across a product's life cycle.
From lab promise to production
A promising material in the laboratory is only the first step. Scaling production means meeting performance expectations, regulatory requirements and financial constraints at the same time. Rao calls that a tightrope walk. His argument is that AI's ability to model multiple variables could help teams handle those competing demands.
Where Rao sees the pieces converging
Looking ahead, Rao expects AI, automation, advanced modelling, digital twins and connected data to come together. In his picture, laboratories become more intelligent and manufacturing plants more efficient. He stresses that no single technology carries the change. The combination matters, and so does the ability to connect processes that generate large amounts of data. He also frames the opportunity as bringing these capabilities together responsibly and at scale.
Affiliate links: we may earn a commission from qualifying purchases, and it never changes our data or rankings. They open Amazon's search results: prices and availability are shown there.