The goal of developmental robotics is to enable an agent to learn
high-level concepts for objects and actions from low-level sensory input
and motor output. Towards this goal we have developed the Qualitative
Learner of Actions and Perception (QLAP).
Initially QLAP motor babbles in its environment. video (1.6 MB)
As QLAP explores its learning becomes more directed.
video (2.1 MB)
This video shows the agent exploring its environment after seven hours
of experience. It has autonomously learned actions to manipulate the block,
and it interacts with the block because this is what it finds interesting at
this point in its development.
After learning, QLAP can be given a task such as hitting a block
in a specified direction.
Here is the agent attempting to perform that task before learning. video (2.4 MB)
Here is the agent attempting to perform that task after learning. video (2.7 MB)