Google DeepMind launches new AI models for robotics
Google DeepMind has introduced two new AI models, Gemini Robotics 1.5 and Gemini Robotics-ER 1.5, aimed at improving robots' understanding and ability to perform complex tasks.

Vad har hänt
Google DeepMind has introduced a new system consisting of two AI models, Gemini Robotics-ER 1.5 and Gemini Robotics 1.5. Gemini Robotics-ER 1.5 acts as a planner and orchestrator, tasked with identifying the necessary steps for a mission. Gemini Robotics 1.5 is responsible for physical execution based on natural language instructions. This system is designed to handle complex multi-step commands for robotics.
Key facts
| Antal modeller | 2 |
|---|---|
| Planerarmodell | Gemini Robotics-ER 1.5 (VLM) |
| Exekveringsmodell | Gemini Robotics 1.5 (VLA) |
| Huvudsaklig funktion | Separation av planering och exekvering |
”this new two-model system will split the planning and execution tasks. Gemini Robotics-ER 1.5 will serve as the planner or orchestrator, determining the steps needed, while Gemini Robotics 1.5 is responsible for executing the physical functions based on natural language instructi”
Varför det spelar roll
The separation of planning and execution addresses previous limitations of single-model systems, which often resulted in errors and delays when attempting to plan and execute actions simultaneously. This division aims to improve robots' ability to perform more advanced and precise tasks by providing a clearer division of labour between thought and action.
Vem påverkas
This affects developers and researchers in robotics, as well as companies using or planning to implement advanced robotic solutions. Users of robotics technology can expect more efficient and competent robots across a range of sectors in the future, including industry and logistics.
EU-status
Not relevant for EU status.
Mer att veta
Initial tests and demonstrations have shown robots sorting items into different bins (compost, recycling, waste) based on instructions, indicating the system's ability to handle complex, multi-step operations.
Quick answers about this story
Vad har hänt?
När hände det?
Varför spelar det roll?
Vilka applikationer har visats?
Länken öppnar i nytt fönster och leder till utgivarens egen sida.
Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.