Rockwell Automation, Inc (NYSE: ROK), a global leader in digital transformation and industrial automation, has announced a major advancement in applying generative artificial intelligence directly at the industrial edge. The company is integrating NVIDIA’s Nemotron Nano technology—an advanced small language model built for low-power, high-efficiency computing—into FactoryTalk® Design Studio™ and future Rockwell workflows. This development brings conversational, reasoning-based AI closer to where decisions are made: on the plant floor and inside operational technology systems.
The collaboration draws on NVIDIA’s open-source Nemotron-Nano-9B-v2 model and the NVIDIA NeMo framework, enabling Rockwell to tailor the AI for secure and performance-critical industrial use cases. Using fine-tuning approaches and training data from FactoryTalk Design Studio Copilot, Rockwell has created an AI model capable of running at the edge without relying on external cloud infrastructure. This design supports environments where uptime, cybersecurity requirements, and physical deployment constraints prohibit traditional AI solutions.
The new model is engineered to work across industrial tasks including design, diagnostics, production monitoring and asset support. It can run on operator panels, dedicated hardware, workstations, or private cloud systems, and is compatible with air-gapped environments. Early testing shows significant gains in accuracy, reasoning capability, and response speed, making the system particularly effective for real-time interaction and guidance in manufacturing environments where latency and reliability are critical.
Tony Carrara, Business Manager for FactoryTalk Design Studio, said the initiative marks a shift in where and how AI can be deployed in operational industries. He noted that by adapting NVIDIA’s model and training it using relevant engineering and automation data, Rockwell is enabling engineers, operators, and maintenance teams to rely on AI support without sacrificing control, predictability, or security.
Joey Conway, Senior Director of Generative AI Software for Enterprise at NVIDIA, stated that the integration demonstrates how small language models can unlock real-time decision-making outside traditional data centers. He emphasized that bringing AI to equipment and smart edge devices is essential for scaling intelligent automation across manufacturing, infrastructure, and industrial systems. Rockwell plans to demonstrate the technology at Automation Fair 2025 in Chicago, November 17–20, where attendees will see how edge-deployed generative AI can streamline engineering tasks, reduce design timelines, and support live operations with context-aware insights.
Rockwell Automation continues to invest in advanced technologies that strengthen autonomous operations and human-machine collaboration. With approximately 26,000 employees serving customers in over 100 countries, the company remains focused on shaping a more intelligent, secure, and sustainable industrial future.
