Once upon a time, learning was a human endeavor—extracting knowledge from books, acquiring insights from teachers, or participating in various workshops. Now, these educational approaches seem to be migrating to the realm of robotics. With technological advancement, scientists are teaching robots to perform tasks by observing humans or watching videos.

Gone are the days when robots were limited to executing simple, repetitive actions in factories or warehouses. Today, they're absorbing skills from more complex domains like cooking or cleaning. All thanks to a revolutionary concept in robotics: teaching machines through observation.

Deepak Pathak, an extraordinary professor at the Carnegie Mellon University's Robotics Institute, embarked on a project that aimed to teach robots tasks through observation. Under his guidance, two machines learned to complete 12 different tasks, ranging from opening oven doors to lifting pots from the stove or picking up items like phones or cans of soup. They did all this by watching videos of humans performing these actions. Pathak explains that the robot can learn where and how humans interact with different objects just by watching videos.

Previous methods of training robots involved humans manually demonstrating tasks or rigorous training in a simulated environment. These methods were not only time-consuming but also prone to errors. This prompted Pathak and his team to propose a new method called WHIRL (Training a Human Imitating Robot in the Wild). WHIRL required humans to perform tasks in the same environment as the robot.

However, the latest research, known as the Vision-Robotics Bridge (VRB), takes this concept even further. Thanks to VRB, robots can learn without human interaction and function in diverse environments. Moreover, a robot can learn a new task in just 25 minutes.

The secret to this new learning method lies in the concept of affordance. Derived from psychology, affordance refers to the possibilities an environment offers to an individual. In the case of VRB, affordances were used to determine where and how a robot can interact with an object based on observed human behavior.

To achieve this, the research team leveraged extensive video datasets, such as Ego4D and Epic Kitchens, portraying everyday human activities. After watching these videos, the robots were able to imitate human actions.

Tasks accomplished by robots trained using the VRB method can be tremendously beneficial in daily chores. Robots could assist in cooking, house cleaning, and various routine tasks. This could significantly enhance people's lives, especially those with mobility or time constraints.

Of course, robotics research is ongoing, but the immense potential of this field is already apparent. Who knows what the future holds? One thing is certain—robotics is ushering in a new era where machines can learn not just from us but from each other.