Abstract
We propose a novel approach to the design, development and implementation
of coordinated classes of autonomous and semi-autonomous taskable vehicles
, both ground based (UGVs) and airborne (UAVs). The agents will be
equipped with a unique cognitive architecture which makes on-line learning and
reconfiguration possible. Hence, the agents will have the ability to respond
to environmental and situational variables and re-task themselves autonomously.
Alternatively, situation information transmitted to the remote warfighter(s)
will make it possible to perform re-tasking remotely.
The system proposed here will enable the warfighters of the future to
control groups of intelligent machines, enabling them to operate as a single
unit in the performance of particular tasks.
Last modified: August 5, 1997
Gaurav Sukhatme, email:
gaurav@robotics.usc.edu