What is Affordance Learning
An approach in robotics in which the system learns to understand what actions are possible with objects and the environment.
Definition
Learning the possibilities of action helps the robot not only recognize an object, but also understand what can be done with it. You can take a cup by the handle, a door can be opened, a button can be pressed, a box can be moved. The system connects the perception of an object with acceptable actions and their results.
Example
A robot in a warehouse sees a box and evaluates which side is safer to grab with its manipulator.
Why it matters
The term is important for practical robotics: recognizing an object by itself does not guarantee that the robot will be able to interact with it correctly.
How it works
The model learns from observations, simulations, or actions of the robot. It relates the shape, material, position and context of an object to the likelihood of a successful action.
Where it is used
- robotic gripper
- home robots
- industrial manipulators
Limitations
The system may make errors with new objects, non-standard materials, poor lighting, or a changed environment.