Time: started 8:11 Thursday, completed 10:47 (interrupted by briefing 9-930).

Weather: low 15c (59F) high of 30.4c (86.7F), hot and dry (5% humidity) with gusting winds

Crew Physical Status: Putting out 100%, like our CPUs. One external support team member was hospitalized in Provo from a heart problem, condition stable; perhaps related to a pre-existing respiratory illness.

Narrative of Field Mission Results: We started with an earlier briefing, at 8 AM, and continued past 9 PM. The entire team worked together persistently and unwaveringly through the day.

Our objective was a morning EVA at Lith Canyon before it became too hot. Twelve of us drove out to the canyon at 8:45 and prepared. Rick's group finished fine-tuning the backbone network at 10:15. Meanwhile, I had the opportunity to meet with Isaac Lopez and Mike Cauley from Glenn Research Center. Sitting in camp chairs in the sun of the rental truck above the canyon, we had a leisurely conversation about center partnership, project management, and research interests.

The first EVA run of the day began about 1020 when the astronaut computers were started. They were connected to generator power, on tables under the canopy. The actual EVA didn't begin until exactly two hours later. Problems we had to resolve included: Too much traffic over the long-distance backbone (commanding computers using VNC from the hab); initializing voice commanding and the biosensors, and starting the Mobile Agents models. Although we had done this many times before, problems can arise from the long-distance communication during initialization, from USB-COM port mapping variations, and other hardware-operating system factors we have yet to isolate.

The EVA proper lasted 40 minutes, with one problem and one notable achievement. For the first time we were testing having the astronauts instruct the ERA to go to a waypoint that was not in the astronaut plan itself. When that didn't work as planned, we used the workaround of having the astronaut carrying the InfoPack (Brent) go to the desired location by the cliff's edge and name that location. Then Abby instructed the EVA to go to that point. This worked! It was our first demonstration of a "shared name space," by which the agents running on different computers (Abby's & the ERA's) could understand a name created by an agent on another computer (Brent's).

The move of the ERA was itself an exciting event to witness. It headed from the area of the truck straight towards the head of the canyon (a huge gaping hole). About 5 meters short of the edge, it was clearly veering too far to the right. Remarkably, the robot went in reverse and backed up almost all the way back to its starting location! Then it moved forward again and this time arrived precisely at the predetermined waypoint.

Meanwhile, Abby and Brent had be standing to the side and didn't get much further, as both of their computers shut down unexpectedly, apparently from overheating. It was now 110 PM and in the upper 80s F, near our operating limit. We shut everything down and most of us returned to the hab for lunch and rest. Over five hours in the desert sun and winds left us all feeling drained and tired.

Besides eating, I used the time to recharge my batteries (the camera ones) and select photos from Tuesday for the web. At 4 PM we decided to return to Lith Canyon for a second evening run. I suggested we retain the HabCom computer at MDRS, rather than running it from the truck at Lith, despite the gusty winds that might endanger the computer network backbone. Bringing the HabCom to Lith would be retreating, a move I thought appropriate for the last day of testing if necessary, but not on a Wednesday with two more days open. Fortunately, this turned out to be a good decision as the winds never caused a problem with the network.

Arriving at Lith just before 5 PM, we found warmer temperatures than at the hab: 89.5F, winds averaging below 10 mph gusting to 17.5 mph, and 5% humidity. We immediately learned that the hab would have a power outage to change generators, a very frustrating, unnecessary event, given that we had people who were dedicated to supporting these utilities. In the warmth of a hot summer afternoon with dust whirling all about, you aren't very happy when people let you down in this way. Thus, on a tight time line, we lost 20 minutes waiting to hear that the hab was back online.

We began setting up the backpack computers at 525 PM, which involved placing the laptops in the backpacks, making all connections (audio, USB), powering them from an external generator, doing a cold restart, rebooting the Mobile Agents "Kaos" directory servers on the Rocky-C ATV (acting as a relay from the far point Cisco repeater on the hill a half kilometer north). Unfortunately, we then experienced another hour of delays, caused by a flawed ERA compilation and a network stack problem (evidently caused by starting the voice commanding process before the network was established). All told, we restarted the astronaut computers three times (at 525 PM, 7 PM, and 721 PM). Fortunately, after declaring this would be the last try, everything synched up, and by 730 PM Abby and Brent were putting on their backpacks.

We then discovered that the Push To Talk headset used by Brent (from the MDRS supply) was stuck in the talk condition. So we wasted 15 more minutes fumbling with it, and resolving to let it go. (In retrospect, I could have made a note last week when something similar occurred during the Pooh's Corner EVA, so we would have fixed this sooner. It's a mark of our needing more support personnel to handle communications.)

The EVA proper, the second run for the day with this scenario, started at 750 PM. Eleven of us had been struggling, sitting, dozing, talking, rushing about for nearly three hours, waiting for this moment. And of course, now the fun begins. The initialization by voice commanding ("Start tracking my biosensors... GPS...") went quickly enough, and they announced the activity of walking to the Fossil Hill. Unfortunately, when the ERA was commanded to move to an appropriate waypoint, so their descent into the canyon could be transmitted by the ERA's video camera, nothing happened. We had fixed the Mobile Agents plan for this EVA to allow the astronauts to refer to waypoints known by the ERA, so why wasn't this command working? We later determined that the CPU load on the ERA computer (one of two) running Brahms was 100%. The command was received by the ERA, but only after we gave up and directed the robot by teleoperation to move.

Abby and Brent proceeded into the canyon. Meanwhile, we set up some flood lights, creating a surreal movie set-brilliant orange lights flared above red metal supports at the canyon head, back-dropped by a solid blue-purple sky. The ERA sat at the lip of the canyon, having announced audibly through its loudspeaker that it had arrived at waypoint 4. From this position, video of Abby and Brent was transmitted back to the hab. Meanwhile, Abby and Brent took photographs and made voice notes, named locations, and associated these. At Rick's suggestion, we had covered the aluminum covers with highly reflective mylar to reduce the absorbed heat; the astronaut's backpacks sparkled as they moved about the canyon floor.

Just as the flood lights were becoming useful, both astronaut backpack computers shut down about 820 PM. The ambient temperature was in the low 80s, and we were all quite comfortable now with the sun below the horizon. But for the computers, the change was only 5-7 degrees F from the day's peak, and apparently the metal boxes, despite the side fan, were serving to hold in the computer's heat. We recalled no problems like this last week, when we ran an EVA with battery swapped for several hours. But then the ambient temperature was much cooler. Clearly we still had an overheating problem.

Despite this somewhat abrupt end to the day's test, it was obviously time to go home, so we quickly disassembled and stashed the computer gear, ERA, tables, and chairs into the truck for safekeeping. We left the site about 9 PM and were back at the hab by 920 PM.

I sent a brief checkin email to mission support, and then turned to preparing a dinner of marinated lemon salmon, rice, and salad. We invited Barney Pell to join us, as he had been taking advantage of our fast internet link on the lower deck to prepare proposal materials for the Exploration program. Dinner conversation was steered away from our work, though on finding that today was Rick's birthday, we soon settled on long discussions about how our productivity changed as we got older, when we got our best ideas, how we wrote papers, and so on. We capped the meal with brownies lit by candles, to which we sang the traditional Happy Birthday. Maarten followed this with the Dutch rendition, which we all enjoyed, especially his ending with a beer bottle waving in air, lyrically promoting a glorious long life.

Plans for Tomorrow: Retest start up procedures; rerun EVA segment #1.

Report Transmission Schedule: GreenHab reports will continue.

Maintenance: LPG supply expected to last 3 weeks was consumed in half that time. We detected this Tuesday and ordered more, but it didn't arrive as hoped. So we switched to diesel power in late afternoon.

EVA Narrative/Data/Interpretations: Pending reports

Inventory: We purchased more vegetables and milk..

Miscellaneous: Dogged persistence may be necessary to achieve difficult goals.