At CES 2026, Hyundai Motor Group is not just showing off robots. It is laying out a vision for how artificial intelligence moves off the screen and into the physical world, and what that shift could mean for how people work, build, and collaborate with machines.
Under the theme Partnering Human Progress, the Group unveiled its human-centered AI Robotics strategy, positioning itself as a key architect of the emerging Physical AI era. This is AI that does not live solely in software or cloud infrastructure, but learns through movement, perception, and interaction in real environments. Factories, warehouses, logistics hubs, and eventually cities become the training ground.
The message is notably optimistic. Rather than framing robotics as a replacement for human labor, Hyundai Motor Group is leaning hard into collaboration. The future it is pitching is one where robots reduce physical strain, absorb risk, and unlock new kinds of work for people, starting where the need is most immediate: manufacturing.
This vision builds directly on the Group’s CES 2022 theme, Expanding Human Reach. Back then, the focus was on robotics hardware and mobility. In 2026, the conversation has evolved. The hardware is still there, but now it is paired with AI systems designed to adapt, learn, and operate autonomously in shared spaces with humans. These robots are no longer static tools. They are becoming interfaces between data, physical environments, and human intent.
At the center of this shift is Boston Dynamics’ Atlas humanoid robot, unveiled in its product form at CES 2026. Atlas is designed for industrial use, not demos. It moves through spaces built for people, handles heavy and repetitive tasks, and is engineered to operate safely alongside human workers. Hyundai Motor Group plans to deploy Atlas at its Metaplant in Georgia beginning in 2028, starting with parts sequencing tasks that are physically demanding and time-consuming for humans.
Over time, Atlas will expand into component assembly and other complex operations, following a careful, process-by-process validation model. Safety and quality come first. The goal is not speed for its own sake, but trust. As Atlas proves itself, deployments will scale across global production sites, gradually reshaping factory floors into environments where humans oversee, train, and collaborate with robots rather than compete with them.
Atlas itself reflects how far humanoid robotics has come. With 56 degrees of freedom, fully rotational joints, and tactile sensing hands, the robot is built for dexterity and precision. Most tasks can be taught in under a day. Atlas operates autonomously from day one, including automatic battery replacement, and can lift up to 110 pounds. It is water resistant, designed for industrial washdowns, and capable of full performance across a wide temperature range. These details matter because they signal something bigger. This is not a prototype chasing a viral moment. It is infrastructure.
Hyundai Motor Group expects humanoids to become the largest segment of the Physical AI market, and it is organizing its entire value chain around that belief. Unlike companies approaching robotics purely from a software or research perspective, Hyundai brings a massive real-world ecosystem with it. Manufacturing plants, logistics networks, component suppliers, software platforms, and decades of production expertise all feed into what the Group calls its end-to-end AI Robotics value chain.
That value chain is the foundation of the Group Value Network, a strategy that connects robotics development directly to real manufacturing environments. Robots train where they will eventually work. Data flows from factory floors into AI systems, improving perception, motion, and decision-making, then back into products and processes in a continuous feedback loop. It is the same philosophy that powers modern software platforms, applied to machines that move, lift, and collaborate.
The Software Defined Factory plays a critical role here. First introduced in Singapore and now evolving at the Metaplant in Georgia, this approach treats manufacturing as a data-driven, adaptive system. Robots operating in these environments continuously learn from real production data, sharing insights across facilities and improving performance over time. The factory becomes both a workplace and a training platform.
To accelerate this learning cycle, Hyundai Motor Group is launching a dedicated Robot Metaplant Application Center in the United States in 2026. Think of it as a gym for robots. Here, human movements like lifting, turning, and recovery are translated into behavioral datasets that robots can learn from and refine. When combined with real-world operational data from factories, the result is a continuous retraining loop that makes robots faster, smarter, and safer with every iteration. Atlas robots trained through this system will begin sequencing tasks by 2028 and advance to complex assembly work by 2030.
Partnerships are central to making this scale. Hyundai Motor Group is deepening its collaboration with NVIDIA, using AI infrastructure, simulation libraries, and development frameworks to accelerate Physical AI innovation. In Korea, this partnership has expanded into a national initiative with the Ministry of Science and ICT, reinforcing the idea that robotics is not just a corporate strategy, but a competitive pillar for future economies.
Within the Group itself, affiliates are aligned around commercialization. Manufacturing infrastructure and production data come from Hyundai Motor Company and Kia. Component innovation, including high-performance actuators, is led by Hyundai Mobis, which is leveraging its automotive expertise to standardize robotics components for scale and cost efficiency. Hyundai Glovis brings logistics optimization to ensure robots can be deployed, serviced, and supported globally. Together, these pieces form a rare, vertically integrated robotics ecosystem.
By 2028, Hyundai Motor Group plans to manufacture up to 30,000 robots annually. That scale matters because it signals a shift from robotics as bespoke hardware to robotics as a platform. To support that transition, the Group is pushing a Robotics as a Service model that treats robots like continuously evolving products rather than one-time purchases. Software updates, remote monitoring, maintenance, and performance optimization are baked into the experience, lowering barriers to adoption and ensuring long-term value.
This model is already in use with global partners across logistics and manufacturing, and it reflects a broader understanding of how technology actually gets used. Robots that improve over time, integrate seamlessly into operations, and deliver measurable returns are far more likely to become foundational infrastructure rather than niche tools.
Hyundai Motor Group is also investing heavily to support this vision at a national and global level. In Korea, the company plans to invest KRW 125.2 trillion over five years starting in 2026, with a strong focus on AI-powered robotics and future growth engines. In the United States, a USD 26 billion investment over four years beginning in 2025 will expand collaboration with leading companies in robotics, AI, and autonomous driving, while establishing a new robotics production hub with an annual capacity of 30,000 units.
Beyond its own ecosystem, Boston Dynamics is pushing the frontier further through a strategic partnership with Google DeepMind. By combining advanced robotics hardware with cutting-edge AI foundation models, the collaboration aims to give humanoid robots greater reasoning, perception, and interaction capabilities. The ambition is not just smarter robots, but robots that can safely and efficiently scale into real, high-impact roles.
On the CES show floor, all of this becomes tangible. Atlas, Spot, and Stretch demonstrate industrial tasks. Autonomous vehicles, charging robots, and collaborative wearables show how Physical AI extends beyond factories into everyday systems. Together, they sketch a future where AI feels less abstract and more present, embedded into environments people already navigate.
Hyundai Motor Group’s CES 2026 narrative is ultimately about confidence. Confidence that AI can move into the physical world responsibly. Confidence that humans and robots can share spaces productively. And confidence that progress does not have to come at the expense of people.
In this vision of the future, robots do not replace human work. They reshape it. They take on the strain, the risk, and the repetition, freeing humans to focus on creativity, oversight, and decision-making. It is a future that feels less like science fiction and more like infrastructure quietly clicking into place.