Environmental Engineering Reference
In-Depth Information
planning-and-scheduling Remote Agent. This agent may also utilize a wide
variety of AI products, including state modeling, case-based reasoning, and
neural nets.
6.2.9 Science Instrument Commanding and Configuration
Other than to support collection of SI diagnostic data and transition into
SI safemode, no SI commanding and configuration functionality needs to re-
side in the FSW backbone. That functionality could be provided by a Re-
mote Agent. The actions of the agent could be driven by receipt of template
structures (generated by the planning and scheduling agent) defining what
SI configuration is needed and/or what operational usage is desired. Separate
agents could be assigned for each SI, or a single Remote Agent could handle
the entire job. The agent(s) would also have responsibility for verifying the
legality of any directives issued to an SI.
6.2.10 Science Instrument Data Storage and Communications
Although management of engineering data storage, as well as management of
transmission of that data and SI diagnostic data, must be the responsibility
of the FSW backbone, storage onboard and transmission to the ground of
primary product SI data collected during a science observation are purely
associated with satisfying science mission objectives and may, therefore, be
entrusted to a Remote Agent. With a lights-out control center, one can easily
conceive of all the duties associated with storing and transmitting SI data
being handled autonomously by Remote Agents, where no general principle
dictates whether these agents belong better in the ground system or the flight
system. Instead, the decision on their location and relative distribution of
responsibilities could, with a generalized flight/ground architecture, be made
on the basis of simple convenience on a mission-by-mission basis.
6.2.11 Science Instrument Data Processing
Other than processing of SI data required to monitor the H&S of the individual
SIs, no SI data processing need be contained within the FSW backbone. The
advantage of assigning this functionality to a Remote Agent is that it enables
cooperative behavior with other agents where the coupling of their functional-
ity can yield a greater whole than the sum of their individual functions acting
in isolation. For example, following collection of wide field data from a CCD
detector in the course of a scan over a region of the celestial sphere, the data
could be processed onboard by this agent, which (using a case-based pattern
recognition algorithm) could identify point sources appropriate for more de-
tailed study. The agent would report its results to the planning-and-scheduling
agent, which would notify the slew agent to execute a return to the specified
target coordinates. At the same time, the SI commanding-and-configuration
