 |
| Google allows bots to perform office maintenance tasks |
Alphabet, the parent company of Google, has announced that Project Daily
Robots (a team within Experimental X Labs dedicated to developing
global learning robots) has moved some of its model machines from the lab to the research giant's Bay Area headquarters for light missions. Implement.
We now have a fleet of more than 100 prototype robots that independently perform a range of useful tasks in our offices,” Hans-Peter Brondmow, director of robotics at the Everyday Robotics Project, said in a blog post. The garbage sorting robot itself can now be equipped with a squeegee to wipe the table, and with the same clutch, teach it to hold the cup and open the door.
These problematic robots are essentially wheel arms with a multifunctional coupling at the end of the flexible arm that connects to the central tower. There is a head on top of the tower with cameras and vision sensors, and the side looks like a rotating lidar unit believed to be used for navigation.
As Brondmo points out, these bots were first seen when Alphabet launched the Everyday Bots project team in 2019. The biggest promise for the company (and many other startups and competitors) is that machine learning will allow bots to operate in rather than structured environments such as homes and offices.
Companies can now make machines that can perform repetitive tasks in factories. But it is difficult for the machine to repeat simple tasks such as cleaning the kitchen or folding clothes
You may have seen a Boston Dynamics robot stumble backwards. But you haven't seen them take out the trash because it's hard to get machines to deal with things they've never seen in the new environment (which humans do every day). And this is exactly the problem that Alphabet wants to solve.
The new rendering is an update on how the company builds standalone devices and tests them in real offices.
Google bots can clean coffee tables
Alphabet's goal is to develop robots that can learn on their own. You don't need to carefully program it to solve a single problem in an environment.
Most robots still operate in structured and tailored environments, and the tasks they perform are very specific. The robot has been carefully programmed to perform these tasks in the right way at the right time. However, this method is not suitable for complex and chaotic rooms in our everyday life.
Robots can transfer learning from the virtual world to the real world. It is now possible to learn complex tasks (such as opening doors) in a single day. This is a higher success rate than it was five years ago. It took the equivalent of four months to register something.
In addition to using machine learning techniques such as reinforcement learning, cooperative learning, and demonstrative learning. Robots are gradually gaining a better understanding of the world around them and improving their daily tasks.
The Everyday Robots project team envisions that in the future machines will work with people to simplify everyday tasks. Google's long-term goal is to enable seniors to work independently for longer.