Google Launches New AI Model For Robotic Industry That Grows Rapidly

JAKARTA Google, a subsidiary of Alphabet, launched two new artificial intelligence (AI) models specifically designed for robotics applications on Wednesday, March 12. The model is based on Gemini 2.0 and is aimed at meeting the needs of the fast-growing robotics industry.

According to industry experts, the robotics sector has made rapid progress in recent years thanks to the increasingly sophisticated development of AI and models. This accelerates the commercialization of robots, especially in the industrial environment.

The launch came a month after robotics startup Figure AI ended its agreement with the OpenAI 'ChatGPT maker' after achieving internal breakthroughs in AI for robots.

Google introduced Gemini Robotics, a vision, language and action-based AI model that allows robots to provide output in the form of physical action.

The second model, Gemini Robotics-ER, is designed to increase robotic understanding of the surrounding environment and enable developers to run their own programs by leveraging Gemini 2.0's reasoning capabilities.

Google mentions that this AI model is compatible with various types of robots, including humanoids and industrial robots used in factories and warehouses.

The use of robotics-specific AI models such as those made by Google and OpenAI can help startups with limited funds to reduce development costs and accelerate the launch of their products to the market.

Google revealed that the Gemini Robotics model has been tested using data from its two-arm robotics platform, ALOHA 2, but can also be adjusted for complex use cases such as Apptronics-made Apollo robots.

Last month, Apptronic raised $ 350 million in funding round led by B Capital and Capital Factory, with participation from Google, to increase production of AI-based humanoid robots.

Previously, Google had acquired a pioneer company of Boston Dynamics robotics in 2013. However, about four years later, the company was sold to SoftBank Group Corp.