This paper, presented at the 8th International Conference on Space Operations, in Montreal, Canada, is available in
PDF
format.
Deep Space Mission Systems (DSMS) lack the capability to provide end to end tracing of mission data products.
These data products are simple products such as telemetry data, processing history, and uplink data. Additionally,
there is a need to create Complex products (packages) which can contain one or more of the simple products, along
with descriptive information about the package itself. The ability to track both types of these data products is crucial
for providing visibility of in-process status along the data product pipeline between space and ground. Other applications
of this information include measuring performance of the space-to-ground data product pipeline and the timeliness of operational
activities, which are key metrics in mission-critical, large-scale embedded systems that exist in multi-node space-to-ground missions.
In order to facilitate this tracing, information about the product format and the observation (metadata) can be defined, captured, and
cataloged per data product in the pipeline. Science teams can then use the cataloged metadata to discover the location and format of data
products cu
rrently in the pipeline to engage in work earlier, and with a better understanding of the data products themselves. There are
certain challenges; however, in producing a system that can meet these requirements, namely (1) The distributed nature of the mission data
products makes it very difficult to catalog and trace along their respective lifecycle in the space to ground pipeline. (2) Both the data
product formats, and the system interfaces which expose the data products are heterogeneous in nature and thus very difficult to integrate.
(3) No clear methodology or standard exists to describe a process for data product packaging, including the format of the package, and what
metadata should be stored about the package itself.
At the Jet Propulsion Laboratory, we are addressing this very problem. We are investigating a distributed computing infrastructure for releasing
science products out of the mission pipeline. We are creating a DSMS Product Service and DSMS Data Packaging Service based on the OODT middleware,
a Data Grid middleware technology that has proven successful in providing access to heterogeneous, disparate data, and federating that data using a
common data model. We seek to provide the capability to (1) Package the distributed data products returned from nodes in the DSMS pipeline (nodes
can be spacecraft, satellite, ground stations, etc.), and create package metadata attached to the product package itself (2) Conform our data product
model to a packaging standard, such as the one proposed by CCSDS and (3) Catalog Products in the DSMS pipeline so that they can be retrieved, and
packaged. Early results from this work suggest that it is a feasible approach to address the challenges described above.