Environmental Engineering Reference
In-Depth Information
statistical analysis of the data, outputs in report form the results of the anal-
ysis, and generates appropriate alerts regarding any identified anomalies. So,
in contrast to an autonomous process, in this case, the ground system per-
forms no independent decision making based on realtime events, and a human
participant is required to respond to the outcome of the activity.
An automated onboard process example could be an attitude determina-
tion function not requiring a priori attitude initialization. The steps in this
process might be as follows. On acquiring stars within a star tracker, an algo-
rithm compares the measured star locations and intensities to reference posi-
tions within a catalog, and identifies the stars. Combining the reference data
and measurements, an algorithm computes the orientation of the star tracker
to the star field. Finally, using the known alignment of the star tracker rel-
ative to the spacecraft, the spacecraft attitude is calculated. The process is
an automated one because no guidance from FOT personnel is required to
select data, perform star identification, or determine attitudes. However, the
process does not define when the process should begin (it computes attitudes
whenever star data are available), simply outputs the result for some other
application to use, and in the event of an anomaly that causes the attitude
determination function to fail, takes no remedial action (it just outputs an
error message). In fact this calculation is so easily automatable that the pro-
cess described, up to the point of incorporating the star tracker alignment
relative to the spacecraft, now can be performed within the star tracker itself,
including compensations for velocity aberration.
On the other hand, the more elaborate process of autonomy is displayed
by a ground software program that independently identifies when commun-
ications with a spacecraft is possible, establishes the link, decides what files
to uplink, uplinks those files, accepts downlinked data from the spacecraft,
validates the downlinked data, requests retransmission as necessary, instructs
the freeing-up of onboard storage as appropriate, and finally archives all vali-
dated data. This would be an example of a fully autonomous ground process
for uplink/downlink.
Similarly, a flight software (FSW) program that (a) monitors all key
spacecraft health and safety (H&S) data, (b) identifies when departures from
acceptable H&S performance has occurred and (c) independently takes any ac-
tion necessary to maintain vehicle H&S, including (as necessary) commanding
the spacecraft to enter a safemode that can be maintained indefinitely without
ground intervention, would be a fully autonomous flight fault detection and
safemode process.
1.2.2 Autonomicity vs. Autonomy
In terms of computer-based systems-design paradigms, autonomy implies
“self-governance” and “self-direction,” whereas autonomic implies “self-
management.” Autonomy is self-governance, requiring the delegation of
responsibility to the system to meet its prescribed goals. Autonomicity is
