ATLANTIS
These agents will not scale well.
Scaling of these
agents involves either the addition of new robotic capabilities or the
application of existing capabilities in more complex ways.
The task queue
and the lack of a mechanism for representing dependencies of operations
forces the agent to perform one operation at a time. Therefore, additional
robotic capabilities will not be fully utilized and without further
extension of the architecture, simultaneous use of current capability is
not supported unless hard-coded as primitive responses. This would probably
still require the extension of the architecture to provided graded or
conditional cognizant failure.
Return to the Top of this architecture.