Agile Robots teams up with Google DeepMind for brainier industrial bots
The collaboration promises adaptable factory bots amid a global race to capture the physical intelligence market.
Agile Robots, a German maker of versatile robotic hardware, has struck a research partnership with Google DeepMind to infuse its machines with advanced artificial intelligence. The collaboration marries DeepMind’s Gemini Robotics foundation models — capable of multimodal reasoning — with Agile’s actuators, sensors and full-stack platforms.
The aim is to create adaptable systems for manufacturing, from electronics assembly to heavy logistics, where robots must handle the unpredictable.
The two firms will exchange real-world data to refine their technologies: Agile’s deployments, numbering over 20,000 units across factories in China, Europe and America, will feed DeepMind’s models, while the AI sharpens Agile’s hardware for tasks demanding vision, touch and planning. Initial applications target precision sectors such as 3C electronics, new energy vehicles, data-centre maintenance and heavy industry. Agile’s existing wares, blending German engineering with mass scalability, now gain a shot at true autonomy.
This pact underscores a pivotal shift in robotics, where disembodied AI meets physical form amid a $200 billion industrial-robotics market strained by labour shortages and geopolitical supply risks. As America’s Figure AI and China’s state-backed humanoids vie for dominance, Europe’s entry via DeepMind, Alphabet’s moonshot arm, hints at a transatlantic counterweight, potentially accelerating the dawn of factories run by thinking machines rather than mere repeaters.
“The huge opportunity ahead lies in autonomous, intelligent production systems that can transform entire industries. Integrating Google DeepMind’s Gemini Robotics models into our robotic solutions positions us at the cutting edge of this rapidly growing market,” Zhaopeng Chen, founder and chief executive of Agile Robots, said in a press release.
Yet factories pose sterner tests than labs: dust, variability and safety rules have dashed prior AI dreams. Agile’s foothold and DeepMind’s prowess in complex reasoning offer grounds for cautious optimism. Carolina Parada, head of robotics at DeepMind, hails the tie-up as “a vital step towards realising embodied AI that learns and adapts like never before”. Success could redefine toil from Shenzhen to Stuttgart.