Complex Environments
in the Meta-Reasoning Architecture
Complex Environments
in the Meta-Reasoning Architecture
MAX encounters two problems when faced with complex environments:
- Slow down
-
MAX must encode a great deal of knowledge to deal with complex
environments. The production-based execution in MAX experiences a
constant factor slow down with an increase in knowledge.
- Perception
-
As with many architectures, MAX was designed independent of its
perceptual systems. That is, it was assumed that the perceptual systems
would convert sensor data into knowledge immediately usable by the
system. In this manner, progress could be made on the general
architecture while delaying consideration of the extremely difficult task
of perception. Before MAX can be integrated into a complex
environment, adequate sensors must be built for that environment.
If adequate perceptual systems can be built, and if the amount of knowledge
needed doesn't cause a fatal slow down, then MAX will function in a complex
environment.
Return to the top of this architecture.
Go to a discussion of this environment
for multiple architectures.
Current Location: Meta-Reasoning - Environments - Complex Environments