Basic collaboration environment:

Continuing last year's pattern, the RST meeting was held over a phone conference, with some members in BuddySpace, and screen sharing using NASA's WebEx virtual meeting server. All of this worked smoothly, with application sharing of particular value as we passed control between RST members, and switched between applications (Science Organizer, Compendium, JPEG and PDF viewers). There is no question that robust screen sharing to allow anything to be shared opportunistically is fundamentally important. One can never predict in advance which tools may be needed.

Compendium:

Compendium [www.CompendiumInstitute.org] is being trialed in two new respects this year, in addition to discussion and idea mapping as conducted last year.

Firstly, as Shannon mentions in her report, the RST was able to construct a detailed EVA plan to pass to the crew for them to review in their planning meeting. The RST walked through a series of templates prepared by Maarten who had briefed Al, filling in the blanks. This worked well as a form of collective modeling. The intention is that the crew (Maarten specifically) will not now have to construct these maps himself based on an informal EVA plan from the RST, but will use them direct. We are waiting to hear how well we did in creating these - in other words, can the Brahms agent read them without error? We wondered next year about the RST having their own Brahms VM running so that they could verify their plan's syntactic correctness themselves.

Secondly, this year's field trial is devoting more attention to the scientific questions motivating data collection on EVAs. Tying the data back to the original scientific rationale is a knowledge management task, and we will be investigating how Compendium can support the movement from data to evidence in relation to hypotheses.

On the 'back office' side of things, Compendium has been stretched to the limit with this year's dataset, and we are looking into performance issues for large deletions, and SQL/XML import/export across platforms. The RST missed the web exports of the data, which we did not manage to complete due to a bug. However, they were able to access the photos and voicenotes from SciOrg, and gained much value from the post-EVA analyses by the astros, in which they collaged photos and added annotations to clarify when a particular voicenote had been taken.

Navigating Science Data:

The RST is deluged with science data as it comes in from the astros, and under time pressure, the challenge is create tools to augment the filtering, visualization, navigation and annotation of data, and the sharing of interpretations. Linking collected photos and voicenotes for the RST ("joining up the astro's eyes and voice") was a key requirement identified in last year's field trial, and doing so automatically is a technical challenge which we are tackling through prototyping the Science Media Player (Southampton Univ.) and the Science Blog (Open Univ.) to give different views of data by time, activity, location, astro, media, science question and possibly other useful indices.

Science Blog:

The Open Univ. has been working on the notion of a web environment for 'distributed scientific journaling', and is investigating the weblog (or 'blog') paradigm and technical infrastructure as a means of delivering this. The idea is that an RST could comment on/argue about collected science data at any time, and be supported in this by an environment which renders the science data as entries in a communal blog, and gives each RST member their own personal science journal (blog), plus views of the collective analysis. Journal entries could be made about an individual photo or a whole cluster of items, and filters would assist by showing, for instance, "all the photos which are inconsistent with hypothesis 3".

A preview of the Science Blog [RST private website] was shown briefly during the meeting when it helped resolve the identity of a photo used by Brent in his analysis. The blog's photo album layout assisted in spotting it quickly (this might be a useful view for the Science Media Player to provide, amongst others).

Shannon saw the potential of the Science Blog for the RST consultants to feed in their analyses via the Web, linked to the photos/voicenotes, and we plan to officially release the blogging tool later today in order to test the system over the remainder of the week.

BuddySpace:

BuddySpace [http://buddyspace.org] has been used heavily by the Open Univ. team (working from the campus and members' homes in UK and US) for joint work in 1-1 mode, and GroupChat mode, but has not yet been used by the rest of the RST.

Hexagon:

The Hexagon video portal was briefly shown during the RST, and excited some interest [http://hexagon.open.ac.uk]. A dedicated room has now been set up for the RST to try, as another channel for maintaining a sense of connection over time and space. Many of us have never met each other face-to-face.

Video Conferencing:

Finally, there was discussion prior to the field trial about the use of internet video conferencing instead of phone conferences, so this may be trialed by members of the Open University team for New York/Milton Keynes collaboration. The OU's 'lite' web-based video conferencing tool [www.FlashMeeting.com] can now be launched direct from a BuddySpace text chat session with a single word command, moving participants from textchat mode to a richer mode of communication. Video conferencing is undeniably a slightly more complex business than phone conferencing in terms of robustness and fluency of audio, unless you use a 'heavyweight' v-c tool like the e-Science Access Grid which has high end audio+video [www.accessgrid.org]. This is a possible candidate to consider in future field trials as the RST infrastructure evolves.

Science Organizer:
Written by Dan Berrios

The ScienceOrganizer Team work this rotation has been particularly exciting, providing great insight into capabilities and limitations of the existing system as it supports the MDRS crew and remote science team activities. We've had several opportunities to study and improve the new Geographic Information Server that automatically creates EVA maps using imagery from Microsoft's TerraServer aerial and topographic image database. These maps list and provide links to collected science data in ScienceOrganizer. We also hope to have an opportunity this week to study the use of the new collaborative image annotation service that ScienceOrganizer now offers. Using this service, up to fifteen users can simultaneously create, view, share, and save written and drawn annotations on any type of image, for reviewing and planning EVAs. Finally, the team is working hard to release new export capabilities during or shortly after the field test, so that crew and remote science team can import data directly into other programs like Excel and other data analysis tools.

Crew/RST Communication and Collaboration:
Written by Shannon M. Rupert

I once worked for a guy who told me "Do something, even if it is wrong." In other words, if I was working alone and needed to get to the next step in an experiment, he didn't want me waiting around for him to supervise me, he just wanted me to take the initiative and move forward, regardless of how things turned out. I find doing remote science much like working back in his lab, because often times we have to "do something" without really knowing if it is the right thing to do.

We waited for much of last week for the crew to go on EVA, worrying along with them when problems cropped up in the field. Then the first EVA was not at all like the EVA we had planned for all those months prior to the field test. In our planning, the first EVA was an ERA Recon EVA. In the end, Brent and Abby went on an astronaut only EVA, and the RST was not involved in the planning for this EVA simply because this change of plans happened quickly and there was no time. The crew saw a chance to complete a successful EVA and took it. It was underway before the RST even knew it was happening.

Reviewing the huge dataset as it rolled into my inbox, one email at a time, was one of the most rewarding things I have done in a while. Last season, we found the staggering amount of data impossible to navigate and even harder to understand. Each scientist was looking for something different in the data. We had no unifying themes to direct our investigations and this made it extremely hard to do any real collaboration, not only between the crew and the RST but even among members of the same team. This season, however, the crew and the RST had done extensive pre-planning together, and had addressed how we might more easily collect and make sense of the data. Now it was time to see what effect, if any, this pre-planinng had on the crew and RST's ability to work together.

I started with the voicenotes. I was struck by two things. First, I could have followed this EVA in "real time" (there is a built in 15 minute delay) if I had wanted. I could get photos and hear details field notes by two scientists out in the field as they were working. And second, I understood what Brent and Abby were doing. Their plan was clear to me and it was easy to seek out and find the data I wanted. This suggests that our idea about pre-planning was spot on in terms of making communication more effective. I think this is particularly striking when you realize that the crew went on EVA without any input at all from the RST, yet they accomplished the goals we had outlined as a joint team. I am not done with the analysis, and have actually only just scratched the surface, but there is a big difference from last night season in terms of communicating the science. I am really excited about comparing this dataset to one from last season in terms of how effectively we were able to do science this season versus the last.

In terms of our investigative themes, the astronauts looked at four questions under our Follow the Water theme and one under the Regolith Terrain Mapping theme during this EVA. They were:

Follow the water:
1. What is the composition of the channels we see?
2. What is the source?
3. What is the geomorphology of the channels and rills?
4. What is the relative age of the different channels?

Regolith terrain mapping:

1. How many different regolith types are in this area?

These questions weren't discussed specifically between the crew and the RST but rather grew out of the collaboration all the scientists had done as a team prior to the field test. The crew's independent work based on the themes exceeded all of my expectations about the effects our pre-planning would have on accomplishing our science goals.

Prior to the SOWG meeting, the crew had informed the RST that the tentative plan was for another astronaut EVA on Tuesday, so they needed the RST to create an EVA plan for them that would follow up on Friday's EVA. We created a plan based on the data and analyses we had received from the crew. We wanted the crew to do two things:

1. Return to the site of the first EVA and take another pan photo, this time from a slightly different location, and use that photo as a basic to continue the regolith terrain mapping. Collect samples as needed.
2. Follow the forked channel in the photo to the source(s) and take coordinates and photos of that area.

We had never actually created an EVA plan using Compendium that could be uploaded into the Brahams agent, and this was our first attempt. Last season, the RST had provided detailed input for EVA planning, but the crew had created the actual plan. This made no sense in an operational sense, because in a real Mars mission, the RST would probably have the responsibility for creating the plan, while the crew on Mars was busy with other things. It was a challenge, but we did it. The crew would still have to correct it, because we had made mistakes in the map, but we had done it!

We finished out SOWG meeting pretty well satisfied with how things were going. And here's where things got really interesting. While we were meeting, Brent has sent a detailed analysis of the regolith mapping that he and Abby had started. He was looking forward to finishing the task that we had just put into the plan! Without any email communication, using only the data we had collected, we had independently come up with the same basic idea for the next EVA.

We learned last season that the crew and the RST have to work asynchronously and sometimes independently of each other, simply because of time constraints. We can't always wait for each other's thoughts and ideas before we act. But it does appear that working together to define broad science goals prior to the field test has softened the challenges of this asynchronous collaboration.