NGK and Laboro.AI Use Generative AI to Cut Simulation Development Effort by Two Thirds

NGK Corporation and Laboro.AI Inc. have built a new way to develop large scale simulation software using generative AI, and they are already putting it to work in NGK’s organic compound crystal discovery service.

At the core, this is about speeding up something that has always been slow and fragile. Simulation software in manufacturing usually depends on manual coding and verification. That means long timelines and a constant risk of errors. What these two companies tested is different. They fed mathematical models, constraints, and conditions into a generative AI in a structured format and let it generate the code.

The result is not just working software, but working software without the usual shortcuts or misinterpretations. With the right rules and base code in place, the AI handled most of the implementation. The human workload dropped to roughly one third, shifting focus back to what actually matters, theory, model design, and interpreting results.

- Advertisement -

Also Read: Mitsubishi Heavy Industries Adopts SmartESG and AI Tool to Tighten Sustainability Reporting

The process is used to predict crystal forms because small changes in conditions affect the solubility and stability of materials which is crucial for pharmaceutical applications. Instead of running endless experiments, simulations can now narrow down viable conditions faster.

Step back and this is bigger than one use case. If this approach scales, it changes how simulation software gets built across industries. Less time writing code, more time solving the actual science.