RST's Crew Analysis & Plan Review from Day One EVA

As everything has a first try, so does collaboration between Mars crews and remote science team back on Earth. Of course, we are not on Mars, but in our simulations we include the fact that the Mars crew will not be having teleconferencing with their science colleagues on Earth. Part of this year's research is testing certain asynchronous science collaboration tools.

Although today's EVA to Pooh's Corner was cancelled due to severe winds, the day started extremely early with the RST having a facilitated meeting over the web, reviewing the crew's analysis and planning of today's EVA. The story goes like this:

As I described in yesterday's ECSR the crew analyzed the panoramas that Boudreaux sent back to the hab in almost real time (see /MDRS/fs03/0427/com.asp). Using tools to view the panoramas, color print-outs of the panoramas and pen to make personalized markings and notes, our two geologists had an inside "field day". We experienced what the MER science team must experience daily, i.e. "what the heck are we looking at and where is it on the map?" After this exciting effort, we had a crew analysis and planning meeting that I facilitated using our meeting capture tool called Compendium (see http://kmi.open.ac.uk/people/sbs/). I videotaped the meeting using my digital video recorder connected to iMovie on my Powerbook standing on a tripod in one corner of the hab. I also captured the screen of the Compendium tool on the laptop I used to capture the meeting (using screen capture software). The screen of the Compendium pc (called the HabCom computer) was displayed on the wall of the hab using a LCD projector. With this setup I could show the panoramas to the crew, capture the discussion in Compendium and the whole bit was digitally captured on disk. I have done this kind of meeting facilitation for years using Compendium, so I would call myself a relative expert in using this tool. Because of the long research history and experience with using Compendium, our effort is not an experiment in the use of Compendium in meeting settings. There is an extensive body of work that has shown this already (not in the least by Al Selvin and Dr. Simon Buckingham-Shum at the Open University in the UK). Coming in to this field test, I thus already knew that in order to use Compendium in this way we needed an experienced Compendium facilitator. Since I am on this crew, voila, no problem on that end.

After our crew meeting ended around 8pm my work was just beginning. I now needed to convert the two digital video files into MPEG-4 format to send to Earth for processing and inclusion in the RST's MeetingReplay tool (more on this later). The meeting took around 45 minutes, and thus to generate the MPEG-4 files it takes as much for Quicktime to create each of the two video files. Luckily, this could be done on each machine and thus it took about an hour to finish this job. This is not an easy process and it seems clear to me now, after having experienced the pain of digital video processing, that these tools still lack significant human-centered design. This should not be this difficult and it is clear that we need to develop better and easier tools to make this process painless for the crew on Mars. How? Well, that is food for future thought, but it suffices to say that there is a long way to go.

I ended up with two enormous MPEG-4 files, one of 105 Mega-bytes and the second one of 257 Mega-bytes. I now had to send these files back to a streaming video server at NASA Ames over our satellite connection from NREN. I don't mean to be disrespectful to our NREN colleagues, because they are doing an incredible job of providing us with 24x7 internet connection, but with plus-minus 200 kilobits per second transmission speed it took well over four hours to transmit these files back to Earth. Nevertheless, with reliability better than my cable-modem at home, our satellite connection was solid and the files were transmitted correctly.

So far so good, but we are only half way (see my future description on the Crew-RST workflow process on the Mission Info page). Now our colleagues from the Intelligence, Agents, Multimedia Group of the Southampton University take over. I e-mail them the semantic export file of the Compendium meeting database. Using this file, and the two downloaded video files they create the MeetingReplay tool (see the RST Communication Systems Report of today, and see also http://www.aktors.org/coakting/mars_replay.html). This tool, which is still created "by hand" by the Southampton team, allows the RST to play back the video of the meeting and the video of the Compendium screen capture synchronized over time.

Figure 1. Meeting Replay Screenshot

Figure 1 shows the MeetingReplay Tool that was created from the video files I downloaded and the Compendium export file I sent to Southampton. This tool is not simply a video replay tool. The two videos (the larger Compendium video and the smaller video of the meeting participants) are semantically synchronized. This is difficult to describe, and the easiest way to explain it is by pointing to the meeting timeline bars at the bottom of the screen capture. This timeline shows all participants. Each participant has a different color timeline. It shows when the participant speaks during the meeting. By rolling over the timeline bars with the mouse, the tool displays the topic related to the Compendium screen that was discussed at that point. By clicking on it, the video jumps to that topic in the meeting. In other words, the user can jump to relevant pieces in the meeting by these semantic links into the content of the discussion. By trying this out yourself, you will experience this impressive piece of software. Go to http://jabber.open.ac.uk/~krp/mdrs_day1/ to try this out.

The MeetingReplay tool is our current answer to the communication time delay problem between Mars and Earth. It is as good as it gets with today's asynchronous collaboration tools. The RST is able to view the meeting with this tool. After this, the RST gets together in a teleconference, facilitated by a Compendium facilitator (Simon has this honor as a Compendium guru), but now using a web-teleconferencing tool. The RST reviewed the crew's plan in Compendium in this web-based facilitated meeting with Compendium, and provided feedback back to crew using the crew's Compendium database send to Earth. Thus, the RST discusses the content of the crew's Compendium map and explanation in the MeetingReplay tool, and provides their feedback within the crew's own context. This makes it easier for the crew to understand the RST's review (see Figure 2).

Figure 2. RST Compendium feedback

The Compendium map shown in Figure 2 is the map created by the crew and sent to the RST, The RST has now annotated their comments into the crew's map. For example, the green plus signs show a critical comment from the RST for the crew. This is what we mean with the RST adding their feedback within the crew's own context. The crew also creates a RST summary map in Compendium, summarizing their feedback to the crew's analysis and plan for the next day (see Figure 3).

Figure 3. RST feedback summary

The last step in this process is that the RST Uplink Lead (Simon again) sends the RST's Compendium database back to the crew in the hab, via e-mail. Since we are on a one EVA per sol cycle, the RST does all this work while the crew is sleeping (in the early hours in California and Arizona). When the crew wakes up (around 8am), the RST has finished their meeting and the Compendium database has been sent back to Mars. The crew's Compendium facilitator (me) brings the RST feedback in the crew's Compendium database and together they view and discuss the RST's feedback in order to create the final EVA plan for the day (see Figure 4). It is by design that we give the crew the final decision on the EVA plan. The reason is that, unlike the MER rovers on Mars, the RST has no direct control over the crew's decision-making process, and ultimately it is the crew deciding what they will do. Figure 4 shows the final EVA plan we created this morning after reviewing the maps from Figure 2 and Figure 3. We printed out the methodology map the RST provided for us, so that the EVA astronauts can follow the RST's science data collection process. Since the EVA was cancelled, because of weather, our geologists on the crew worked part of the day on refining the RST's data collection method provided in their feedback. We hope to report on this in the coming days, but I can already say that our tools provided us with the ability to have asynchronous collaboration between a distributed Earth-based RST and a crew on Mars. I hope you can tell that I am already a happy camper! After only three days into the field test we have already successfully tested two-thirds of our capabilities. Over the next week we hope to do this several times and improve our abilities, but I can already say that this has been an unbelievable experience.

Figure 4. The crew's final second day EVA plan

A report from a happy, although tired, crew member ... till tomorrow. Signing off!