Subject: Re: abstract for AIAA Space 2000 Date: Fri, 14 Apr 2000 13:40:13 -0700 From: "Bryan A. Biegel" Organization: NASA Ames Research Center To: Al Globus Al, Regarding your AIAA abstract, it looks pretty good. I thought that too much text was spent on your favorite topics: RLVs and launch costs. :-) Maybe that's the slant you negotiated with Tony Gross, but I think it would be more satisfying to the audience (and therefore to Harry McD) to spend your time on the important issues (challenges and opportunities) of a solar-system-wide IPG, rather than focusing on launch costs and RLVs. For example, launch cost and safety are much more tied to activities here on Earth than to an extraterrestrial IPG. It seems that you are trying to morph this talk into something that falls under your proposed RLV work, and that doesn't seem the best approach for the target audience. Make sure to point out that one of the goals of the IPG is to integrate one-of-a-kind sensors (e.g., space probes, satellites) for easy access in an ISE (integrated synthesys environment). This is also one fo the challenges, not only because of the unique data and access methods that must be hidden by IPG interfaces, but also because of these spacecraft are transient - they start life at some point, they end life at some point, they may be inaccessible for periods in between, and (as you mentioned), they may have large access latency. IPG services must facillitate the clean attachment and detachment of instruments. For example, an Earth-based "mirror-site" of the instrument's status and data will probably be accessed most of the time, in the same way that many companies are out-sourcing mirror sites to allow surfers to get faster web access. Putting space probes on the IPG seems analogous to putting vending machines on the Web. This had two main benefits. It allows people to access their state, to see what they could get from the machine at any given moment. Secondly, it allowed software agents to check the state of the machine, and to request service when necessary. Similarly, you don't have to have a fully programmable computer in a spacecraft. If software agents running on a "real" computer (probably on Earth) can read the state of the craft, they can doo all of the necessary analysis, and then alert humans or even direct the craft automatically. If the craft *does* have its own intelligence, but perhaps limited data storage, it can use the IPG/Web to pull information as needed to decide what to do, rather than only being told what to do. By the way, in my view, the IPG will be just a set of access-restricted HPC and special instrument areas of the Web. These restricted areas might be called ISEs. Thus, extending the IPG to the solar system to me is just interfacing astronauts, satellites, RLVs, space probes, etc. to these ISEs. I could be wrong about what the IGP is, but I believe that in order for the IPG to be a success, it must leverage the Web, not try to compte with it and use other technologies and hardware. Anyway, a couple other minor comments on the abstract.... > NASA Ames and her partners I would avoid using gender-specific pronouns where possible. It isn't "PC", and (more importantly), it distracts from your message. In this case, "its" will work fine. Mentioning that better IT systems could improve mission safety and success while reducing cost was a good idea. As you must know, this was a major finding of the Mars98 failure analysis team. My personal opinion is to avoid discussion of disabled people in space. Rather than showing what the IPG will enable, this sounds to me like a boondoggle of trying to ship physically disabled people to space (where, by the way, people become more disabled over time). It's my personal viewpoint. Anyway, good work so far. -Bryan