AivexaNewsSearch
AI news for builders and product teamsChecked every hour

Innovating with AI in the special chemicals industry

Collected Nov 9, 2026

Lubrizol, a specialty chemicals manufacturer, is building a digital foundation that ties together business processes, modern data platforms and cloud infrastructure to support its research and development teams. Girish Rao, the company's vice president of information technology, described the effort in a webcast produced in partnership with Infosys, in content published by MIT Technology Review's custom content arm rather than its editorial staff. The stated aim is to give researchers tools to test ideas faster, make better-informed decisions and develop more sustainable materials without giving up performance.

The core of the approach is combining decades of accumulated technical data with the expertise of chemists and engineers who understand both customer needs and the physical constraints of the work. Data scientists sit alongside them to spot patterns, test hypotheses and model possibilities before anyone runs a physical experiment. Rao framed the intent narrowly: the technology is meant to give scientists better tools, not to replace their scientific judgment.

That framing points at a real tension in industrial R&D. Rao described three groups approaching the same problem from different angles: scientists prioritize experimental validity, data scientists prioritize model performance, and IT prioritizes platform security and scalability. Lubrizol addresses this by organizing the groups into what it calls fusion teams, aligned around shared objectives and starting with small pilots to build trust before scaling. Rao said that once team members trust one another and understand the value they are creating, progress improves considerably.

Cloud infrastructure underpins the strategy by giving the company access to computing resources and modern tooling without the substantial capital and maintenance burden of building its own data centers. For a mid-sized industrial firm, that is a common trade-off: renting elastic compute and managed data services lowers the upfront cost of experimentation, while introducing questions about data egress, vendor dependence and where sensitive formulation data physically resides. Rao did not detail which cloud provider or platforms Lubrizol uses.

Sustainability is where the company sees particular value. Digital modeling lets researchers evaluate more formulations, explore alternatives such as starch and corn-based materials, and weigh trade-offs across a product's full life cycle. This is a meaningful shift in method: instead of synthesizing and testing candidates one at a time, teams can screen a much wider space of possibilities computationally and reserve bench work for the most promising leads. A likely trade-off, in my reading, is that model output is only as good as the historical data behind it, and decades-old experimental records are often inconsistent in format and conditions.

Rao was careful to note that a promising lab result is only the first step. Scaling production while meeting performance expectations, regulatory requirements and financial constraints remains difficult, and he suggested that AI's ability to model many variables at once could help teams balance those competing demands. This is the classic valley between discovery and commercialization in chemicals: a formulation that works in a beaker may fail on cost, fail a regulatory threshold, or fail to behave the same at plant scale.

Looking ahead, Rao expects AI, automation, advanced modeling, digital twins and connected data to converge, with laboratories becoming more intelligent and manufacturing plants more efficient. His emphasis was on integration rather than any single tool, arguing that connecting these processes is what will generate large volumes of data. Digital twins, for context, are software replicas of physical assets or processes that let engineers simulate changes before applying them; in chemicals they are typically used to tune process parameters and reduce costly trial runs.

Why it matters: For teams in industrial and scientific software, this is a case study in adoption rather than a product launch, and it suggests that organizational design, meaning shared goals and small pilots across science, data and IT, may matter as much as the models themselves. Developers building for lab and manufacturing environments should expect demand for tools that respect scientific judgment and fit existing expert workflows. A reasonable inference is that the hardest problems will be data integration and validation, not model capability, and that sustainability reporting could become a durable driver of demand for simulation and life-cycle modeling.

Read at MIT Technology Review · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

As companies look to accelerate innovation with artificial intelligence, cloud computing, and advanced modeling, the primary challenge is connecting new technologies to reliable data, scientific expertise, and the practical realities of bringing products to market. For specialty chemicals manufacturer Lubrizol, digital transformation is all about giving researchers the tools to test ideas faster, make better-informed…