Space Task Force Burlinghame Minutes (Sept 2000) ( http://cgi.omg.org/cgi-bin/doc?space/2000-09-05 ) SSGS Minutes ( http://cgi.omg.org/cgi-bin/doc?space/99-11-01 Chairs : Bill Branch, CSC, wbranch@csc.com Jim Wetherbee, Altair Aerospace Corporation, jimw@altaira.com For those of you with little time, I have highlighted some of the activities and then provided some detail content below. Highights : The SSGS has become a Task Force !!! SSGS Overview presentation – http://cgi.omg.org/cgi-bin/doc?space/2000-09-01 CORBA Objects in Space: Advantages and Challenges – http://cgi.omg.org/cgi- bin/doc?space/2000-09-02 Work on a Telemetric RFP CCSDS SLE presentation – http://cgi.omg.org/cgi-bin/doc?space/2000-09-03 SOTG presentation – http://cgi.omg.org/cgi-bin/doc?space/2000-09-04 OMG RFP Process – Evan Wallace of the Manufacturing Task Force SuperMOCA architecture The role of monitor and control as basis for ALL aspects in space Space Day – an information day at the December OMG meeting (visit to Kennedy?) Details : The Space Satellite and Ground System Working group met Wed Sept 13th. There were a number of presentations delivered, some discussion on a Telemetric RFP, as well as some discussion of a Space Information Day to be conducted at the Orlando meeting in December. The meeting commenced with a round-the-room introductions, then Jim Wetherbee presented an overview of the group ( http://cgi.omg.org/cgi-bin/doc?space/2000-09-01 ), where it been, and where it thinks it is going. The group is attempting to become a Task Force at the DTC meeting. A very quick discussion of the SSGS OMG Green Paper, its structure and the organization and description of some of the identified work areas. It was well noted that there is a good deal of work necessary to update the Green Paper. Tim Canham from JPL / MDS presented CORBA Objects in Space: Advantages and Challenges (http://cgi.omg.org/cgi-bin/doc?space/2000-09-02 ). The presentation is based on work done in the MultiMission Data Systems, an architecture focusing on reuse of components, both for ground and flight components. The presentation focused on two different problems, one for the development and simulation phase, the other looking on using CORBA in the flight software. CORBA had many benefits for the simulation phase : 1) the IPC does not have to be developed, 2) language and platform independence (the MDS developed on SunSolaris and deployed using VxWorks), 3) ease of distributability (all on one processor or on many processes without recompilation), 4) an open architecture (just have to understand the IDL and semantics). However, there were some challenges as well including 1) speed – 98% of their time were spent on the IP stack, 2) there is a learning curve (mostly for distributed vs monolithic technologies), 3) with the TAO Orb product, there are versioning challenges as well. For flight software, there were many more challenges including 1) size of ORB product (2MB in the case of TAO, although one ORB vendor stated that they can perform the ORB services in less than 100K), 2) the culture of flight software groups is anti-distributed and their metrics are focused on monolithic programming. There was also discussion about the different uses of CORBA in space – just between objects on board (definitely use CORBA), space-to- space (if close enough, could use "normal" CORBA implementations), and space-to-ground (whose challenges would include a modification to the normal CORBA protocol to include "connectionless" communications). Overall a very informative presentation on real-world use of CORBA in space and flight software. The second part of the meeting focused on trying to get some handles on some requirements for a Telemetric RFP. To this end, Mike Stoloff from JPL gave a discussion of the CCSDS SLE Service and some of the lessons learned from some of their work ( http://cgi.omg.org/cgi-bin/doc?space/2000-09-03 ). The CCSDS Space Link Extension (SLE) is concentrated on the distribution and management of telemetric and commanding processes. Some interesting information resulted from this work. The SLE team separated the distribution of the data for the management services necessary to set up and configure the relationship. This is an important issue that is addressed later in these minutes. The SLE focused on data products consisting of return frames, channel and virtual channels, and packet data. Some of their experiences included a fully asynchronous message-based solution, although there was discussion that this decision could be reviewed in a CORBA-based environment. One interesting point that was made that should be kept in mind when creating an RFP was that it is necessary to ensure that messages are delivered in sequence without duplication. Other things to keep in mind are QoS, as well as the realization that CORBA can help in some areas, however, there may be areas where CORBA is not necessary. Jim Wetherbee then presented some lessons learned from work done in the SOTG, focused mainly on the Data Acquisition area of the SOTG (http://cgi.omg.org/cgi-bin/doc?space/2000-09-04 ). The Space Object Technology Group CORBA-based standard has been implemented in the areas of Data Acquisition and Commanding and has controlled a live satellite through the use of two SOTG-compliant COTS products. There were many lessons learned that should be used to ensure RFP submitter address, or could be used as evaluation criteria for RFPs. The SOTG architecture strived for consistency of mechanisms and methods. Some of the main points are perform publish and subscribe consistently, fetching of data should be able to be performed synchronously and asynchronously, large data sets should be accommodated (with particular care given to performance, distribution, and blocking of clients and servers), use exceptions, have the ability to transport small data structures with the minimum data necessary for performance considerations, although also allow the ability to fetch metadata for the objects. One more point to be careful of is allowing too much flexibility of interfaces can make development much more complicated and problematic, as well as creating possible impedence issues when trying to connect implemented products together. Evan Wallace of the Manufacturing Task Force then gave a brief talk on RFP preparation, generation, and tips for working within the OMG process. Some tips were to watch the schedule and ensure that vendors are prepared to keep the schedule. Specify issues and evaluation criteria well to ensure that the submitters understand exactly what you want. Also investigate the scope of the RFP. Don't try for too much, but also don't keep it too small. Peter Shames then presented some work from the SuperMOCA project, an attempt at an space applications architecture that existed from approximately 1994 – 1998. The architecture focused on monitor and control, especially focused on control loop. Separation of decision logic from the main monitor and control structure was another important aspect of this architecture. This presentation prompted discussion on whether an RFP on objects that support monitor and control functions in a fairly generic fashion should be attempted first, because this concept will permeate all aspects of satellite applications, and interfaces can be created to support these functions, then all space applications objects could be have a common basis and true plug and play could be ensured. We resolved to look at this issue in more detail, however continue along with a Telemetric RFP as well. After this there was talk about a Space Day in Orlando, organized with the OMG (thanks, Majel). The group is planning an Information Day on Tuesday of the next OMG meeting (Dec 12th). The focus will be on the introduction of this group to the community and the OMG. The group is also planning a field trip to Kennedy Space Center for Wednesday afternoon. More information will be provided by mid- October. Anyone interested please contact Bill or Jim. There is plenty of work expected between now and the December meeting. Work on the OMG Green paper, work on a Telemetric RFP, investigation of a monitor and control RFP, and preparation for Space Day in December will keep us very busy for the next few months. Anyone interested in contributing, please contact Bill or Jim. Other notes are that the liaison relationship between the OMG and the CCSDS has been made official. Peter Shames has been named the official liaison. The group has also become a Task Force and is planning for increased activity over the next couple of meetings.