Environmental Engineering Reference
In-Depth Information
of the different spacecraft-level agents I1-I4 determine the organizational ar-
chitecture. The top-down coordination architecture includes only one single
(highly intelligent) I1 spacecraft-level agent, and the other spacecraft are (un-
intelligent) I4 agents. The centralized coordination architecture requires at
least local planning and possibly interaction capabilities between spacecraft,
requiring I3 or I2 agents. The distributed coordination architecture consists
of several parallel hierarchical decision-making structures, each of which is
“commanded” by an I1 intelligent spacecraft-level agent. In the case of a fully
distributed coordination architecture, each spacecraft in the organization rep-
resents an I1 spacecraft-level agent, resulting in a totally “flat organization.”
9.6 Grand View
The next level of space-based autonomy is to develop and verify agents and
agent-communities concepts to the point that they can migrate to actual
ground-based operations and, when fully verified and validated in an op-
erational context, migrate to the spacecraft to provide onboard autonomy.
Figure
9.5
is a view, a grand view (not the only one), of what such a system
might look like [
178
,
179
,
184
], and is one possible representation of
progressive
autonomy
. It paints a picture in which we can see many threads of agent-based
activity, both ground-based and space-based. The major theme of the figure
is that of agent migration from one level to another. The figure depicts the
various migration paths that could be taken by agents and communities of
agents enroute to a spacecraft. This is the essential theme of our proposed
approach to realizing complete autonomy for constellations as well as other
mission types.
Progressive autonomy refers to the levels of autonomy that can be incre-
mentally achieved by a dynamic community of agents. Achieving a higher
level of autonomy in a community means either increasing an already existing
agent's capabilities through reprogramming it, introducing a new agent into
the community with the desired capabilities, or allowing an agent to develop
a new or modified capability via learning.
Progressive autonomy is advantageous for at least two reasons:
1.
It allows a new capability to appear from a community of agents supporting
an operational mission after that capability has been verified and is trusted
outside the testing environment.
2.
A qualified agent can be dispatched to a community in need on a tem-
porary basis. Once the need has been fulfilled, the agent can be removed.
This keeps the operational resource requirements for the community to a
minimum.
Figure
9.5
illustrates some of the concepts that are associated with pro-
gressive autonomy in agent-based communities on the ground and in space.
