Gemini Robotics 2: Google DeepMind Adds Whole-Body Control

What happened

Google DeepMind has announced Gemini Robotics 2, an updated AI model that extends robotic control from the upper body of a humanoid robot to its entire physical form. Where the previous iteration of Gemini Robotics was limited to upper-body manipulation, the new model supports coordinated whole-body motion — from feet to fingertips — enabling humanoid robots to perform far more complex, physically integrated tasks.

The announcement, made on Thursday, marks a meaningful step in DeepMind's push to make general-purpose humanoid robots viable outside of controlled laboratory settings. By giving a single AI model command over a robot's complete kinematic chain, DeepMind is moving closer to machines that can navigate, manipulate and interact within environments designed for human bodies.

Why it matters

For customer experience and service design, the arrival of whole-body humanoid control is not an abstract engineering milestone — it is a signal about the near-term shape of physical service delivery. Retail, hospitality, logistics and healthcare operators have long faced a hard constraint: automation works well in structured, repetitive environments but breaks down wherever a task requires the kind of fluid, full-body coordination that humans take for granted. A robot that can crouch to retrieve an item from a low shelf, carry it across an uneven floor and hand it to a customer with appropriate dexterity begins to close that gap.

From a behavioural economics perspective, the more capable a robot becomes at mimicking human physical competence, the more acutely the question of social presence arises. Customers calibrate trust, warmth and perceived effort based on how an agent moves and responds — not just what it says. Service designers deploying humanoid robots will need to think carefully about motion choreography and interaction rituals, not merely task completion rates.

The Renascence take

The industry conversation around AI in service tends to fixate on language — chatbots, voice assistants, large language models. Gemini Robotics 2 is a reminder that the next frontier of service experience is embodied, not just conversational. Most operators are not yet asking the right questions about what whole-body robotic presence will do to customer perception, staff dynamics and service ritual design.

The risk is that organisations treat humanoid robots as faster, tireless staff — optimising for throughput and missing the deeper behavioural reality. Customers read bodies, not just outputs: posture, gait, hesitation and gesture all carry social meaning that shapes trust and satisfaction. The operators who will lead are those who design the experience of the robot — its movement vocabulary, its proxemics, its failure behaviours — with the same rigour they apply to human service standards. Whole-body control does not automatically produce whole-body hospitality; that still requires intentional service design.