More humanoid robots doing complex jobs.

This time it's the BMW company's turn.

Yes I know, in recent posts I have perhaps talked too much about humanoid robots, but it is exactly what the world is producing at an accelerated rate and understanding what is happening helps us understand the possible future that is approaching.

The BMW Group announced a new pilot project at its factory in Lapesig, Germany, where it is testing humanoid robots called Aeon. The idea is ambitious, using human-shaped robots to perform heavy and repetitive tasks within automotive production. The AEON was developed in collaboration with the Swiss company Hexagon Robotics and was designed to act mainly in the assembly of high-voltage batteries used in electric vehicles.

These tasks require precision, strength and constant repetition, exactly the type of work where theft can have a great advantage. The project is part of a larger BMW strategy called Physical AI, this concept seeks to connect digital artificial intelligence with machines that operate in the real world. Instead of just analyzing data, AI can directly control robots and physical systems within factories.

Before reaching the production floor, Aeon underwent intensive training in virtual environments using simulation platforms at Nvidia such as ISIM, the robot learned locomotion, navigation and object manipulation skills in digital worlds that replicate the real factory. This simulation training drastically reduces the time needed to prepare robots, which previously could take months, can now be done in a few weeks.

In the real world, AEON uses an advanced set of sensors and cameras, that information is processed in real time by computers embedded in the robot based on the Jetsonoring chip, allowing the machine to make decisions and execute tasks accurately. The system can also connect to factory digital twins.

If the tests continue with good results, as until now, BMW plans to expand the use of this type of robot to different areas of production and tell me, have you wondered to what extent these machines can replace more complex human tasks? Some startups are already trying to accelerate this transition much faster than we imagined.