Next: The Digital Zenith Camera TZK2-D - A Modern High-Precision Geodetic Instrument for Automatic Geographic Positioning in Real-Time
Up: Telescopes and Observatory Operations
Previous: The Subaru Telescope Software Trinity System
Table of Contents -
Subject Index -
Author Index -
Search -
PS reprint -
PDF reprint
Koratkar, A., Grosvenor, S., Jones, J., & Wolf, K. 2003, in ASP Conf. Ser., Vol. 295 Astronomical Data Analysis Software and Systems XII, eds. H. E. Payne, R. I. Jedrzejewski, & R. N.
Hook (San Francisco: ASP), 152
Science Goal Driven Observing
Anuradha Koratkar
Space Telescope Science Institute
Sandy Grosvenor
Booz Allen Hamilton
Jeremy Jones
NASA/Goddard Space Flight Center
Karl Wolf
Aquilent
Abstract:
In the coming decade, we will be forced to automate many of the
scientific tasks that are done manually today because observatories
will have to be managed in a fiscally tight environment. Thus,
spacecraft autonomy will become a part of mission operations. In such
an environment, observing campaigns of inherently variable targets and
targets of opportunity will need flexible scheduling to focus observing
time and data download on exposures that are scientifically interesting
and useful. The ability to quickly recognize and react to such events by
re-prioritizing the observing schedule will be an essential characteristic
for maximizing scientific returns from the observatory.
The science goal monitoring (SGM) system is a proof-of-concept effort
to address these challenges. The SGM will have an interface to help
capture higher level science goals from the scientists and translate
them into a flexible observing strategy that SGM can execute and
monitor. We are developing an interactive distributed system that
will use on-board processing and storage combined with event-driven
interfaces with ground-based processing and operations, to enable fast
re-prioritization of observing schedules, and to minimize time spent on
non-optimized observations.
In the last fifteen years astronomers have come to depend on software
tools to operate observatories/missions, and also to obtain, analyze,
archive and reanalyze data. In the coming decade due to limited funding,
observatories must work with smaller and smaller operations staff, even
as instrument complexities and data volumes are increasing. Many of
the scientific tasks that have traditionally been manually overseen
will have to be automated. In such an environment, software tools are
an essential component for increasing scientific productivity and
developing cost-effective techniques for obtaining and assimilating data.
Thus, spacecraft autonomy will become an even greater part of mission
operations. While recent missions have made great strides in the ability
to autonomously monitor and react to changing health and physical status
of spacecraft, little progress has been made in responding quickly to
science driven events.
The current spacecraft operations are reasonable for simple pointed
observations. For observations of inherently variable targets, monitoring
projects and targets of opportunity, the ability to recognize early
if an observation will meet the science goals, and react accordingly,
can have a major positive impact on the overall scientific returns of
an observatory and on its operational costs. Thus, for at least a class
of targets there is a need for flexible scheduling to capture quality
data to achieve science goals.
The Science Goal Monitor (SGM) project is a research effort funded by
NASA Code R to determine if it is feasible to ``get the eye back to the
telescope" and to develop prototype software that will enable this.
In the first phase of our study, we have determined that there is a
subset of space science problems that are conducive to the philosophy of
science goal driven observing. These problems are related to objects
that are time variable and are often monitored for long periods.
Using these problems as test cases for our prototype, we are in the
process of designing and developing the SGM system which is a set of
tools that have the ability to capture the underlying science goals of
an observation, translate them into a machine interpretable format, and
then autonomously recognize and react in a timely fashion when goals
are met. SGM will provide astronomers with visual tools to capture
their scientific goals in terms of measurable objectives and be able to
autonomously monitor the data stream in near-real time to see if these
goals are being met. Our prototype is designed for use in a distributed
environment where some analysis can be performed onboard a spacecraft,
while other analyses can be performed on the ground.
To testbed our proof-of-concept prototype, we have determined that an
observatory that has the following conditions would get the most benefit
from an SGM-like system: a typical target is intrinsically variable,
targets are monitored for long periods of time, and
spacecraft are designed for scheduling flexibility.
The Science Goal
Monitor (SGM) system will interact with not only the data processing
pipeline for a mission, but when used on board the spacecraft, it will
also interact with the raw data from the detector. Figure 1 shows
the high-level concept of the SGM. At present, we are concentrating on
designing the Science Goal Capture Tool (SGCT). We hope to design and
develop an underlying architecture and framework for the SGM system
(see Figure 1). The three key modules include:
Figure 1:
SGM System Overview (ground-based)
 |
- The Science Goal Capture Tool (SGCT) will
capture the user's science goals in an intuitive and
easily understandable manner while simultaneously storing them in a
format for easy machine processing. To achieve this mix of capabilities
we are modeling the SGCT after ``state diagrams" commonly used in computer
science. An example of an early design overview of the SGCT interface
is shown in Figure 2.
- The Science Goal Analyzer (SGA) will interpret the science data
stream from the observatory and interact with its processing
pipeline to analyze whether or not a set of science goals are being
satisfied. When goals are met, the SGA will fire events notifying any
registered observer of the event and provide access to the details.
While usability will be the key in the capture tool, the analyzer
will place a premium on efficiency and speed. We anticipate that the
SGA will work in cooperation with an onboard scheduler. The SGA will
also monitor messages and data from the spacecraft flight
software and update the status of its monitored campaigns accordingly.
This partnership will allow the SGM to introduce progressive autonomy and
dynamic behavior while the instrument and flight software continue to
provide their traditional safety and control checks.
- The Science Goal Monitor Console (SGMC) will provide a
visual interface and console for mission operations to query, monitor,
and interact with the SGA. We currently anticipate that this feature will
be used primarily as a testing module to observe and validate the workings
of the SGA and also as the ``reference implementation" of an event monitor.
An important component of the SGM will be the ability to send alerts
to interested scientists and operations managers, alerting them when
SGM has recognized an interesting event. There are already several event
notification systems in development or operation. We propose to provide
an interface to one or more of these systems, but do not plan to develop
one specifically for the SGM.
We are currently
prototyping user interfaces to capture science goals in a fashion that
the scientist can use and understand. We are also evaluating existing
and emerging software to dynamically evaluate science data on board the
spacecraft. The prototype will be used to evaluate the effectiveness
of an SGM and to understand the risks involved for such a system to
be implemented. We plan to implement and test the prototype using the
Small and Medium ApertuRe TelescopeS (SMARTS) project.
Figure 2:
SGM Science Goal Capture Tool
 |
Introduction of flexible scheduling and autonomously reacting
to science driven events inherently infuses automation technologies
into mission operations. Missions, especially complex high-profile
missions, are more culturally and politically averse to risk when it
comes to automation. Clearly, the capture of science goals rather than
the mechanics of an observation while developing observing programs,
and the subsequent automatic analysis of the data stream to determine if
goals are met, represents not just a leap forward in automation, but a
large change in the operations paradigm. We are in the early stages of
the Science Goal Monitor project to develop prototypes, evaluate their
effectiveness, and understand the risks.
Acknowledgments
This work is funded by NASA Code R under the Computing, Information and
Communication Technologies (CICT) program.
© Copyright 2003 Astronomical Society of the Pacific, 390 Ashton Avenue, San Francisco, California 94112, USA
Next: The Digital Zenith Camera TZK2-D - A Modern High-Precision Geodetic Instrument for Automatic Geographic Positioning in Real-Time
Up: Telescopes and Observatory Operations
Previous: The Subaru Telescope Software Trinity System
Table of Contents -
Subject Index -
Author Index -
Search -
PS reprint -
PDF reprint