Developmental robotics is the use of human psychology principles in the design of intelligent, autonomous robots that learn from their own individual experiences. This field is also known as epigenetic robotics.
Developmental robotics is concerned with the progress of each individual machine's capabilities, rather than on the evolution of robotic technology in general (evolutionary robotics). The most noteworthy features follow.
- At each moment in time, the robot determines its next action based on its current actions and external circumstances.
- The robot acquires new skills based on the skills it has acquired, thereby building on its own abilities.
- The robot builds its knowledge base as it encounters new environments and experiences.
- The robot learns from its own mistakes.
- At all moments in time, the robot deals with signals that come directly from its own sensors.
- The robot learns to cope with new situations by trial-and-error, in much the same way as humans do.
- The robot interacts effectively with humans, as well as with other robots.