Time: 19:00

Weather: low 35.2F (1.8C); high about 50F (~10C)

Crew Physical Status: Very good for a Saturday, energized by the progress and cooler weather.

Brief Narrative of Field Mission Results: After a thirty minute briefing at 9:00, we proceeded to set up the first outdoor test. This process took almost two hours, as we now placed the minibook computers in their plastic containers mounted on top of the Mars Society suit backpacks. In this configuration, a third computer is required to view the desktops of the minibooks and to initialize their processes. One of the Rocky ATVs ran yet another computer, which provided the link between the minibooks, the monitor computer, and the HabCom computer. Sekou meanwhile established that the PDAs in the geologist-astronauts' pockets were receiving biosensor data from the biovest computers worn by the geologists. To coordinate, we used two radio bands, and recorded all transactions using an audio mixer on the hab's upper deck. The scenario is trivial: The astronauts verbally request the Mobile Agents system to start up, and then announce a series of activities: suiting up, leaving the hab, and returning. They then shut down the system. On a second trial just before lunch, the entire process executed smoothly and surprisingly quickly.

After lunch we tested scenario 1B, which involved putting on the helmets with the backpacks and walking on a predetermined route around the hab. This fully scripted EVA took about 1.5 hours and generated a great deal of data. Changes from 1A2: ERA was actually running (though up on blocks in its truck because of snow flurries and mist in the wind); alerts and email were enabled. Successes in the scenario included: Logging a sample bag, making a voice annotation and playing it back, processing of biosensor data with an appropriate high temperature alert, and telling the astronaut (automatically) where he is located.

During the 1B test about a half dozen questions were raised that will require analysis first thing Monday morning. In addition, to avoid complications from false positive commands (i.e., the system recognizes a command that is not actually given), the dialog manager will be modified to ask for confirmation for every command before it is carried out. (Voice recognition will improve through an automated statistical method; currently false positives can be expected to be over 10%.) This will be somewhat tedious for the geologist-astronauts, but will eliminate a potential problem during this early testing stage.

We determined that the wireless network was not introducing delays (latencies < 10ms). When Astronaut 1 moved to the far side of a hill behind the flag, his computer dropped out of the network, but rejoined the Mobile Agents system later without difficulty. Many questions were raise about CPU load, queuing, antivirus software, slow hard drives, etc. Some easy reconfigurations are possible and will be tried Monday.

Interesting initialization issues have been identified and require changes in the start of EVA script: The astronauts cannot tell the Mobile Agents system that they are starting to don their suit because they do not have their headsets connected to their backpacks at that time. Also, they cannot start tracking of their location until they leave the hab because their GPS units have not yet acquired a signal. These are the kinds of interactions that our trials are ideal for discovering.

We are intrigued by the significant improvement on the second run of 1A1 without any software or hardware changes. Hypotheses include: Overhead of starting Windows on the first run; camera drivers not loaded until the camera is connected the first time; GPS signals were locked and retained after the first run.

Successes for the day include: Completion of the first (scripted) EVAs supported by the wireless Mobile Agents system. Feedback to astronauts was better (e.g., they can now give the command "Acknowledge" to be sure that the system is listening to them). First execution of an EVA involving walking over an hour around the hab.

The executive committee met for an hour about 17:30, for an exhaustive review of all functionalities and pending changes.

Plans for Tomorrow & the Day After: Sunday will be our first day off in two weeks (and for some of us, three weeks). Monday we are planning to start by analyzing today's data, and everything else being equal, we will re-run 1B without changing the system. We could also make modifications to handle open problems. We are inclined not to do another bench test, but we will not do EVA #2, which involves a repeater (and a possible delay in the network) until we have tracked down the current handful of puzzles.

Report Transmission Schedule: We will select, prepare, and transmit today's photos either later tonight, early tomorrow morning, or Monday morning.

Maintenance: Frank checked the generator oil; it's down a half quart. Will be filled tomorrow.

EVA Narrative/Data/Interpretations: The scrilpted EVA involved walking to a rock near the radio telescope antenna, then going to the hay, the sprinkler, and observatory. We will include at least one photo taken during this EVA. It was the coldest since we arrived (about 40F), which was ironic given that the team has worked indoors during the warm sun for an entire week.

Inventory: We are amazed to discover that we've eaten about six boxes of cereal in a week.

Miscellaneous: We're ready to go outside for some real exploration, and so look forward eagerly to our day off.