Social RoboticsGoogle DeepMind's Gemini Robotics 2 adds whole-body humanoid control
The new vision-language-action model drove Apptronik's Apollo 2 from spoken instructions and supports robot teamwork.
Google DeepMind released Gemini Robotics 2, a vision-language-action model able to control an entire humanoid rather than just its arms. In demonstrations, an Apptronik Apollo 2 walked to a table, picked up a watering can and placed it on a low shelf after a spoken request, and used a dexterous hand to tie knots and seal bags. A companion reasoning model, Gemini Robotics ER 2, interprets instructions, talks with people and plans tasks, and is available to developers, while the VLA and an on-device version are limited to early partners. Language-driven control of this kind is central to robots that can take everyday requests from the people around them.
Source: The Robot Report
This archive entry summarises reporting published elsewhere; follow the source link for the full story. The summary was drafted with AI from that reporting.