NVIDIA and LG-Doosan: The Alliance That Will Redefine AI Factories

What happened?

NVIDIA has announced two simultaneous strategic collaborations: with South Korean conglomerate LG Group and with Doosan Group, also from South Korea. The common goal is to build 'AI factories' that integrate accelerated computing infrastructure, robotics platforms, digital twins, and synthetic data generation to drive physical artificial intelligence (Physical AI).

In the case of LG, the alliance covers LG Electronics, LG Innotek, and LG CNS. NVIDIA will provide its full-stack AI factory platform, including NVIDIA Isaac Sim, Isaac Lab, GR00T (language model for robots), and Cosmos (foundational world models). LG will contribute its expertise in consumer electronics, home robotics (such as the CLOiD robot), optical components, and cloud services. LG's AI factory will generate high-quality training data for robotics and industrial projects, using NVIDIA Cosmos for data synthesis.

With Doosan, the collaboration spans Doosan Robotics, Doosan Bobcat, Doosan Enerbility, and Doosan Corporation Electro-Materials. Technologies such as NVIDIA Isaac Sim, Newton physics engine, and Jetson Thor will be integrated to advance Doosan's robotic operating system, Agentic Robot OS. Additionally, use cases in industrial tasks like depalletizing and sanding will be explored, and new robotic form factors will be developed, including dual arms and humanoid platforms. Doosan Bobcat also plans to incorporate physical AI into construction, agriculture, and material handling machinery.

Why is it important?

These alliances mark a milestone in the convergence of artificial intelligence and manufacturing. Traditionally, factories have relied on rigid automation; now, with physical AI, robots and production systems are expected to perceive, reason, and act in dynamic environments, learning from simulation and synthetic data. This reduces the need for costly and hard-to-obtain real-world data and accelerates the deployment of collaborative and autonomous robots.

Moreover, the collaboration with LG and Doosan positions NVIDIA as the technological orchestrator of the next generation of smart factories. By standardizing its platform (Isaac, Cosmos, GR00T, MGX, DSX) across large industrial conglomerates, NVIDIA creates a closed but scalable ecosystem that could become the de facto standard for AI-based manufacturing. This has geopolitical implications: South Korea, with its strong industrial base, becomes a laboratory for the factory of the future, while NVIDIA consolidates its dominance in AI hardware and software.

Consequences for the market and users

For the manufacturing industry: A reduction in costs and time for implementing intelligent robotics is expected. SMEs could benefit from preconfigured platforms and synthetic data, although the initial investment in NVIDIA infrastructure remains high. Standardization could fragment the market if other players (like AMD or Intel) fail to secure similar alliances.

For the robotics sector: Companies like Doosan Robotics will evolve from robotic arm suppliers to full AI solution providers. The integration of language models (GR00T) into home and industrial robots will open new applications in logistics, home care, and construction.

For workers: Intelligent automation could displace repetitive jobs but also create new categories (AI supervisors, digital twin engineers). Training in digital skills will be crucial.

For investors: NVIDIA reinforces its position as an AI leader beyond data centers. LG and Doosan stocks could benefit from the 'tech halo.' However, executing these alliances at scale will take years, and competition (such as Tesla with Optimus) is not standing still.

What should readers know?

• These AI factories are not traditional physical plants but ecosystems integrating simulation, training, and continuous deployment of AI models for manufacturing.
• Synthetic data generation via NVIDIA Cosmos is key to overcoming the shortage of training data in robotics.
• LG and Doosan are not mere customers; they are partners co-developing technologies and setting standards. LG, for example, will build a 'physical AI data factory' to serve other Korean and global companies.
• The concept of 'digital twin' extends to the entire supply chain, from raw materials to customer delivery, connected in real time by AI.
• The collaboration with Doosan includes autonomous heavy machinery (Bobcat), expanding the scope of physical AI beyond light manufacturing.

"We are moving from programmed automation to intelligent automation that learns and adapts. NVIDIA is building the operating system for the factories of the future." — Analyst at TheVortiq

Context and outlook

These alliances are announced at a time when generative AI saturates the market and companies seek concrete applications in the physical world. NVIDIA had already presented its vision of 'physical AI' at GTC 2024, and now it materializes that vision with heavyweight industrial partners. The choice of South Korea is no coincidence: the country has an advanced manufacturing infrastructure and strong investment in robotics (it has the highest density of industrial robots in the world, according to the IFR).

Competition is also intensifying: Tesla advances with its Optimus robot, Google DeepMind researches world models, and startups like Covariant develop AI for robotics. However, NVIDIA's advantage lies in its comprehensive platform (hardware + software + data) and its ability to attract large conglomerates.

In summary, the collaborations with LG and Doosan represent a decisive step toward standardizing AI-driven factories, with NVIDIA at the core. The coming months will be critical to see concrete implementation and initial results in productivity and cost reduction.