NVIDIA and Doosan Group Collaborate to Advance Physical AI and AI Factory Infrastructure

Why It Matters

The alliance accelerates the rollout of AI‑driven automation and energy‑efficient data‑center infrastructure, giving both firms a competitive edge in the rapidly expanding AI‑factory market.

Key Takeaways

• NVIDIA's physical AI stack integrated into Doosan Robotics' Agentic Robot OS.
• Doosan Bobcat to embed AI models for autonomous construction equipment.
• Doosan Enerbility explores low‑carbon power for NVIDIA AI factories.
• Electro‑Materials supplies high‑performance CCL for NVIDIA MGX AI servers.
• Collaboration targets scalable AI infrastructure across robotics, power, and data‑center hardware.

Pulse Analysis

The NVIDIA‑Doosan partnership reflects a broader industry shift toward end‑to‑end AI factories, where hardware, software, and power converge under a single ecosystem. NVIDIA brings its accelerated computing platforms—DSX for AI‑factory orchestration, MGX for modular server design, and the Isaac suite for robotics simulation—while Doosan contributes deep expertise in industrial automation, construction machinery, and energy generation. By aligning their roadmaps, the two companies can co‑develop reference designs that shorten time‑to‑market for AI‑enabled equipment, a critical advantage as manufacturers race to embed intelligence in every production step.

In robotics, Doosan Robotics is leveraging NVIDIA Isaac Sim, Isaac Lab, and the Jetson Thor edge processor to evolve its Agentic Robot OS. This integration enables simulation‑to‑real workflows, physics‑based calibration, and on‑device inference, allowing collaborative robots to perceive and adapt to dynamic factory floors. Use cases such as depalletizing, sanding, and dual‑arm or humanoid platforms illustrate how AI can transform traditional robot arms into versatile, task‑specific agents. The move also positions Doosan to compete with global AI‑robotics leaders by offering a full‑stack, AI‑first solution rather than isolated hardware.

Beyond the shop floor, Doosan Enerbility and Electro‑Materials are addressing the power and hardware backbone of AI data centers. Enerbility’s exploration of gas turbines, steam turbines, and small modular reactors aims to supply reliable, low‑carbon electricity for high‑density AI workloads, while its fuel‑cell technology could provide on‑site, clean backup power. Simultaneously, Electro‑Materials’ high‑performance copper‑clad laminates support NVIDIA MGX’s high‑bandwidth server designs, ensuring signal integrity at scale. Together, these efforts create a holistic AI‑factory value chain that spans robotics, autonomous equipment, power generation, and advanced PCB materials, setting a new standard for integrated AI infrastructure.