Google recently announced a new artificial intelligence model that can help robots understand tasks such as throwing garbage. According to Google’s blog post, the Robotics Transformer 2 (RT-2) will be a vision-language-action (VLA) model trained on information and images taken from the internet that can be turned into actions for the robot.Artificial intelligence will train robots
While a task like collecting garbage may seem simple to humans, it requires an understanding of a series of tasks for a robot to learn. For example, the robot needs to first recognize which objects create garbage, and then know how to pick it up and throw it away. Rather than programming a robot to perform these specific tasks, the RT-2 allows the robot to use information from the web to help it understand how to complete the task, even if it has not been explicitly trained in the exact steps.Google said that the new model almost doubles the performance of robots in unprecedented scenarios compared to the previous version of the model. Google added that the new version can use primitive reasoning to respond to user commands. On the other hand, Google does not have short-term plans to widely launch or sell robots with this technology. However, the trend shows that robots will be used more widely in warehouses or as home assistants.
60504
