Saturday, 09-02-06

Three trucks arrived from JSC, with the Command Trailer, the Garage Trailer, and a covered truck with general field gear. The JSC diesel generator and forklift were delivered.

Once the trucks were unloaded in the field, the site was racked and the patio structure was assembled.

Significant thunderstorms were observed 10 miles to the East and West, but no rain fell over the test site. A major dust storm occurred in late afternoon, with 30+ MPH sustained winds for about an hour. Winds were from the South.

Sunday, 09-03-06

Four trucks arrived from JSC, with Scout , the ACES Van, Scout's long conex box, the short conex box, the transport trailer, the Kawasaki Mule, and Centaur. Two Trucks arrived from JPL, with the Pressurized Rover Compartment (PRC), and two ATHLETE robots. The JPL command bus, COACH, and logistical vans and trailers also arrived. The ARC truck arrived, and the GRC satellite truck arrived. A mobile crane arrived for unloading the conex box's from their trailers.

JSC's Command Patio was covered, the Command and Garage Trailers were positioned, the outfitting of the trailers was completed and they were connected to the generator. The two Scout conex box's were lifted into position with a crane. Scout's tents were assembled and positioned with anchors. The ACES van and generators were positioned and setup. The JPL COACH was positioned, and connected to power. The generator was powered up and used for the bulk of the day. The ATHLETES dismounted from their truck, and were checked out from COACH. The Centaur was un-crated, and checked out in the Garage Trailer. The K-10 Orange and K-10 Blue were checked out, and placed in the Garage Trailer.

Cloudy morning cleared in the afternoon. A minor dust storm erupted with intermittent, ~15 MPH winds from the East.

JPL Bus

K10 Robot

Monday, 09-04-06

The GRC generator was delivered, and the network control trailer arrived from KSC.

During the morning session, the robots continued their individual checkouts. The PRC was lifted with the tripod and then placed on one of the ATHLETEs using the LaRC lifting procedure. K-10 Blue was tested in the field and its cameras were calibrated. Scout was tested with GPS way-point collection, and the Kawasaki Mule was used to collect GPs points ~1 mile from camp. Centaur was driven in the field about a mile over the course of the day, testing drive train gains for the soft dirt, rocks and clumps of brush.

Mid day the test leads had a conference and selected the test zone, looking for a flat, horizontal area with little vegetation that would be damaged. The site zone was selected and marked.

During the afternoon session, a complete run of the coordinated field test scenario was completed. An RF interference problem was found on ATHLETE- initial testing suggests it is only near the SCOUT antennae tower, and the robot worked well in the test zone. The ATHLETE carrying the PRC was positioned in the field and lowered onto its hard points. SCOUT was driven into the test zone by shirt sleeve operators, who dismounted and walked to the PRC. Centaur successfully approached and grasped the EVA box from the SCOUT equipment tray. K-10 Blue performed a 360 inspection of SCOUT, and then SCOUT drove-out in an autonomous mode.

Cloudless Day, with a light wind from the Southeast.

Athlete

SCOUT vehicle

Tuesday, 09-05-06

SCOUT performed long range field testing, in a selected region ~1 mile to the West of camp. This test will be repeated with suited subjects next week. Operations went well using shirt sleeve subjects. SCOUT then participated in multiple test runs of the coordinated field test.

K-10 Blue practiced its inspection task on the parked SCOUT, circum-navigating the vehicle and providing photos. The ARC field crew is now fully integrated into the JSC command trailer for operations, and is using the JSC garage trailer for servicing.

Centaur practiced its box removal task, successfully picking up the tool box from SCOUT's equipment tray using a dual arm, grasp reflex. Base driving has been tuned for the soft dirt and is performing well on the mix of sand, rocks, and sparse vegetation.

The LaRC crew ascent ramp was added to ATHLETE A, parked with the PRC on its back. The robot's legs were posed to avoid contact with the ramp. ATHLETE B was readied for some hill climbing tests.

Cloudless morning, with sparse clouds appearing mid day. There was a light wind from the SouthEast. A "dust devil" erupted over the GRC satellite truck, in the lee of a hill. The miniature tornado was about 20 feet in diameter. It rapidly moved across the test site and went directly over Centaur. No damage occurred. ARC personnel had a chair holding their test plan papers overturned. The papers were quickly lifted to 500 ft, and lost to the desert. Two smaller "dust devils" sprang up around camp during the afternoon, one hitting the garage trailer, and the other loitering on the generator. No damage occurred.

K-10 at Camp

Athlete

Left to Right: JPL Bus, Generators, Centaur Command Trailer & Garage Trailer w/ Patio between, SCOUT Garage

Wednesday, 09-06-06

Only morning runs were performed, due to a thunderstorm in the afternoon.

The GRC satellite truck and KSC firewall are now fully operational. Satellite phones have been routed to the command trailers, so that cell phones will not be needed as much.

SCOUT performed long range field testing in the morning, at the site ~1 mile to the West of camp. Teleoperation was tested, using the ACES van. Planning software was tested, and connection to the suit CAI packs was tested. A bolt on the left rear suspension "A-Arm" came loose, and was fixed in the field with a jack.

Centaur completed the last of its task checkouts, with obstacle avoidance, an autonomous driving approach to SCOUT, and an autonomous grasp. Multiple runs were performed, and tasks times were shortened to about 7 minutes. In the afternoon, Centaur was tested in the Garage trailer performing grasping trials, while the weather was bad.

K-10 Blue continued to practice its inspection task, performing a 360 drive around SCOUT. The robot is performing well in the rougher terrain around the SCOUT parking location.

ATHLETE A (with PRC) had to be tarped due to the significant wind and rain. ATHLETE B was used in a test with its bucket positioned in an inverted pose like a backhoe. The robot performed digging in the sun back soil, finding the soil is quite tough more than a couple centimeters deep. The teeth on the bucket dug into the hard pack to full depth, and the high strength of the legs overpowered the soil. As the storm approached, ATHLETE B was also tarped. After the thunderstorm, ATHLETE B was checked out and found to be healthy.

The day start with a clear sky. Mid day a large thunderstorm formed to the NorthWest, and gradually widened and moved over the test site. Sustained ~ 20MPH winds and light rain forced a termination of the afternoon runs.

Centaur Grasps a Sample Box

Satellite Dish

Athlete Module Covered for Rain

Thursday, 09-07-06

Individual CFD practice runs were completed by both Centaur and K-10. The Centaur crew gained more experience with obstacles that approximate planetary camera stands. During all morning runs, Centaur video images and telemetry were continuously shipped to the DRL via the GRC Satellite truck. K-10 added a large rock to the CFD test area to demonstrate its obstacle avoidance.

SCOUT successfully tele-operated the Planetary Exploration Geophone System (PEGS) from ACES. PEGS was also tested using SCOUT's autonomous precision driving capabilities (worked very well). Other items that were tested and functioned exceptionally well, were: 1) Power management system, 2) Mobile Agents Software, and 3) ExPOC voice commanding.

ATHLETE continued to dig using its miniature back hoe.

Two integrated CFD practice runs were completed. SCOUT personnel filled in for suited subjects. Both runs were successful, and techniques to reduce overall run time are being discussed within the groups. The weather rapidly deteriorated during the start of a third run and the run was cancelled. With a strong possibility of lightning in the area, all equipment was secured and the camp was closed down in late afternoon.

K-10 Leaves Garage Trailer

Command Trailer Interior

Patio Platform Between Command & Garage Trailers

Friday, 09-08-06

AM:

The SCOUT team walked through test plan 5 in preparation for afternoon activities.

ATHLETE manipulated a box that approximates a simulated lunar nuclear generator.

The Centaur teamed continued to refine obstacle avoidance and practiced approaching and removing the sample box. The K-10 team implemented a new procedure, reducing inspection time from 25 minutes to 14 minutes. The PRC fan power system was moved to eliminator generator fumes from entering the crew compartment. Donning stands were installed in the PRC and suit technicians were trained in PRC emergency egress procedures. With permission from LaRC and RATS safety folks, the hand rails were removed from the PRC ramp. Brain Wilcox/JPL personally started the removal process with a saw in hand.

After a detailed pre-test briefing with all team members, Centaur, SCOUT, K-10, ATHLETE and Desert RATS team members participated together in their first official shirt sleeve run of the complete CFD. All systems performed well. We even allowed a Cai pack to be worn and it did not cause any disruptive EMI.

David Jerrera (SP) EVA Systems/Constellation Office/HQ came by with Joe Kosmo and we gave him a well received tour of all robots and trailers.

PM.

While preparations were underway for the first suited run of the trip, a storm started to roll in at 1:30 PM. All equipment was secured and team members moved to the Meteor Crater Visitor Center to wait out the storm which included lightning strikes within 2 miles of the test area. The team returned to the test site at 3:30 PM. The test area was fine.

Our first full up CFD suit run went well in spite of the weather. All robots successfully preformed their tasks and for the first time in the desert a suited subject entered the PRC. After the run, the participants were strategically positioned for a CFD family photo. The suited subject is Drew Feustel/CB.

Centaur spent the remainder of the afternoon practicing driving and mapping out non-CFD desert activities for the weekend.

Scout modified TestPlan #5 (due to time constrained - it was getting late and Drew had been in the suit for a long time). Test Plan #5 included PEGS operations by a suited subject (the PEGS project is headed up by Drew). The system worked flawlessly and it was very impressive to have a suited subject driving SCOUT and operating a payload with very little shirtsleeve assistance (only assistance was telling him where to stop - he needs to stop every 3 meters). He was able to activate all the button on the DCS to mode motion, drive with the joystick, monitor a Tablet PC onboard that was collecting seismic information, and push all the buttons on the "Thumper program" that Rob Hirsh wrote.

After over 2 hours in the suit, Drew drove back to suit camp, got off SCOUT, and took the suit off. Overall everyone (Mobile Agents Folks, Cai Pack team, Suit Team, SCOUT team, etc.) was very happy with the test carried out by Drew.

Saturday, 09-09-06

Throughout the day, SCOUT tested fully autonomous operations of the thumper payload. This integration between the Planetary Exploration Geophone System (PEGS) H/W, SCOUT’s onboard systems and Mobile Agents allows for the autonomous acquisition of seismic data. In the past, a person would have to very precisely manually drive to a certain location, another person would manually operate the thumper, and then another person would have to prepare and coordinate the movement to the next thump location. This sequence is typically carried out 25 times during a PEGS operation. Using these new capabilities successfully tested today, the system is now able to performed a series of move, thump, prepare the thumper to move with the press of ONE button. Additionally, this action can be activated with a voice command issued by a suited subject wearing a CAI pack, a local operator in the control trailer or a remote operator in ExPOC. This capability will be demonstrated from ExPOC this coming week.

During the morning, testing of the SCOUT obstacle avoidance continued. This activity will continue throughout the coming week. For periods during both the morning and afternoon, the SCOUT vehicle was made available to Centaur and K-10 to allow fine-tuning for the coordinated field demonstration.

In the morning, Centaur tested operations on uneven, rocky terrain. These teleoperated tests involved driving up and across the hill behind the camp, collecting rock samples and placing the samples into a sample box.

In the afternoon, the Centaur and ATHLETE performed joint activities. In the test ATHLETE deployed an anchor and Centaur tethered ATHLETE to the anchor. This type of operation might be required prior to ATHLETE descending into a lunar crater. Time permitting; a similar activity will be performed during the upcoming week where ATHLETE descends a hill, after being tethered an anchor by Centaur.

Throughout the day, Centaur practiced operations for the coordinated field demonstration. Operations were initiated from both the command trailer in Arizona and from the telepresence cockpit in the Dexterous Robotics Lab in Houston.

The K-10 rover spent the day transitioning from the first week crew to the second week crew and fine tuning for the coordinated field demonstration

ATHLETE spent the morning driving an auger to anchor a simulated lunar nuclear generator.

Robonaut Picks Up a Rock Sample

Robonaut's Centaur Base Navigates Rough Terrain

Centaur Assists Athlete With a Tether Hook

Sunday, 09-10-06

Skeleton crew only, with most of the field team taking a (well deserved) day off.

There was a personnel change over on many of the robot operations squads, with people from ARC, JPL and JSC taking some time to go over status and do a crew "hand off".

SCOUT and Centaur were used in a series of practice runs of the coordinated field test scenario. The work was primarily on the Centaur driving approach to the Scout equipment tray. Multiple control modes were tested, ranging from teleoperation with local obstacle avoidance, to autonomous driving with supervision from Houston.

The day had a continuous string of small thunderstorms across the valley, and required the robots to be "safed" periodically.

Monday, 09-11-06

SCOUT Performed a shirt sleeve run of its Test Plan 4. This test includes the suited subjects driving the SCOUT vehicle performing human following and gesture control. The Thumper was controlled from ExPOC at JSC, in two autonomous modes. Additional SCOUT autonomies were tested. In the afternoon, SCOUT participated in practice runs of the Coordinated Field Test with the other robots, and the second suited run.

Centaur did final testing of all its autonomous control modes. Lunar-like "rocks" were used as obstacles for the long approach to SCOUT, with Centaur being teleoperated while having a local control mode that prevented it from hitting or coming close to obstacles over which it could not drive. For this test the teleoperator intentionally drove directly at the rocks. At high speeds (~10 KPH), the robot attempted to veer away from the rocks, found that it could not, and stopped safely. This allowed the operator to back up, and navigate around the object. At medium speeds (~5 KPH) the robot autonomously modified the teleoperator's commands, driving around the rock with a safe margin. As Centaur approached the SCOUT vehicle, it switched to a fully autonomous control mode, rendezvousing with the equipment tray on SCOUT's rear bumper using stereo vision. Under time delayed supervision from Houston, Centaur then grasped the sample box, using tactical and force guided grasp reflexes.

ATHLETE B Continued its hill climbing work on the terrain behind camp. Control gains were adjusted for the mix of soft and hard soils, to refine autonomous load and chassis leveling. The robot was tested on a scenario securing a box (emulating a nuclear or other power system), using anchoring screws.

The day had a continuous string of small thunderstorms in the morning, with one longer delay in the afternoon. Weather moved from West to East.

Suited Runs with SCOUT

Athlete Climbs a Hill

Tuesday, 09-12-06

SCOUT Performed a suited run of its Test Plan 4. The following functions were fully tested:

1) Suit subjects manually driving with the use of the onboard display and control system.
2) Voice Recognition and Voice synthesis,
3) Autonomous Power Management System,
4) Path Planning,
5) Human Following and Gesture Recognition using the vision system,
6) Mobile Agents Software,
7) Autonomous pt-2-pt navigation,
8) Automated Thumper Operations (from ExPOC and from Base camp), and
9) Cai Pack

Two Scout team members provided interviews for the Distance Learning Network (DLN) with Heather Paul and Debbie Sharp. The vehicle video was reconfigured so that the DLN would have access to video coming directly from SCOUT, which was in turn broadcast nationally. In the afternoon, SCOUT was used in a suited run with the Coordinated Field Test.

Centaur climbed the hills behind camp during the morning, getting on top of the highest hill and setting up for a joint test with ATHLETE. The Centaur base ran at ~5 KPH over rough terrain, on rolling hills, out running engineers trotting alongside. In the afternoon, Centaur returned to the front of camp for a successful run of the Coordinated Field Test, supervised from Houston across time delay.

ATHLETE B climbed to the near the crest of the hill behind camp, taking the very hardest possible path. This included some tall steps up capstone, and the steepest section of the terrain. The operations made a transistion between control modes as the terrain steepened. For the first half, it rolled short distances (10-30cm) on six wheels, stopping to level the chassis and redistributing its weight across the six legs evenly. ATHLETE rolled up approximately 50% of the hill's height. As the terrain became too steep for this mode, it transitioned to an autonomous 3-legged gait, lifting three legs while advancing the body on the other three, then alternating legs. This walking mode was used to move to approximately 75% of the hill's height. Nearing the capstone, ATHLETE transitioned to single leg moves, with each footfall designated over extremely rough steps in the rock. By nightfall, ATHLETE was at 90% of the hill's height.

The day had a continuous string of small thunderstorms in the morning, with two longer delays in the afternoon. Most weather was moving from the SouthEast to NorthWest.

Wednesday, 09-13-06

This day was devoted to final checkouts for the Coordinated Field Test. SCOUT, K-10, Centaur, ATHLETE A with its PRC and the two suited subjects performed multiple test runs. When the team seemed bored, we ran again. And again. Then one more time to be sure.

All systems performed well, and the team is well positioned to identify and handle the few abnormal things that are inevitable in a complex system. Annomalies such as periodic reductions in the network bandwidth are now easily identified, and the team does not waste time chasing loose strings, since we have solutions that are well proven. Switching control modes, taking 10 second pauses for network traffic to clear, or re-ordering events are tactics that have been tested, and are at the team's fingertips. During the tedium of repeated trials, the operations have been refined and the task's timeline has quickened.

The camp has begun to collect a large number of still and video cameras on tripods, and this day was spent refining how to position these cameras for best effect. The number of people following the systems was reduced, so only those required for safety were allowed in the field. The humans and robots are ready.

In the evening, the team decided to take a risk. Centaur was driven back up on the hill and a new task was attempted with ATHLETE B. The task was to drive an anchor into the capstone, set a belay, and then allow ATHLETE to rappel down the hill. ATHLETE B had made it to the top of the hill in the morning during the coordinated field test, and had identified an ideal placement of the anchor. The following sequence was choreographed:

1) Centaur would approach ATHLETE and extract the EVA tether hook, backing up while pulling the rappel line out of ATHLETE's body.
2) ATHLETE Would extracted a screw from its tool holster, position it, drive it into the rock, let go of the tool leaving an eyelet positioned for Centaur, and then lift that limb out of the way.
3) Centaur would drive back into position, connect the tetherhook to the eyelet on the screw, release the tether, and back out of the way.
4) ATHLETE would then used the tethered, screw anchored, line to rappel back down the hillside.

For a first attempt, the task went extremely well, taking about 40 minutes. In the tag up following the test, a couple ideas were agreed upon for improvements, such as how to place the tether hook initially on ATHLETE, and how to pose ATHLETE's limbs to provide better reach access for Centaur. It was agreed to show this task to guests the following day, weather permitting.

Relatively stormy day, with several stops for small showers. The team could see the storms coming for 20-40 miles, and make good choices about when to safe the camp. Lightning strikes were seen in the distance (>40 miles). Wind gusts before each storm were ~25 MPH.

Centaur and Athlete Work From a Hill Top

Centaur Secures a Safety Line for Athlete's Descent

Thursday, 09-14-06

On our last day in the desert, we set aside time for showing our guests several of our robot and space suit activities. As luck would have it, the weather was the worst we had seen in 10 days.

About 12 guests from the NASA ESMD Advanced Capabilities Division at HQ and the LaRC ETDP Program Office were met in Flagstaff and directed out to the test site. Along the way, the group traveled through a large thunderstorm that proceeded East, and arrived at the camp about 30 minutes after the people arrived. In that time, the guests had a chance to receive a briefing from Amy Ross, the lead for the advanced space suit testing, Craig Bernard, the lead for the space suit information packs, and Frank Delgado, the lead for the SCOUT rover. While the robots and people were poised to run through the Coordinated Field Test, the weather broke open and the guests were taken into the Command Trailer and Patio area for shelter. What followed was the most intense rain shower we had seen in the 10 days of field work.

As we waited, hoping for the weather to clear, presentations were given to the guests on the Patio's back wall screens, used normally for displaying operational status. Jaret Mathews (JPL) presented videos of various tests that had already been conducted by the two ATHLETE systems, including moving the PRC, climbing hills, and manipulating drills and buckets with its limbs. Susan Lee and Terry Fong (ARC) presented videos of post processed images from the K-10 vehicles, showing its ability to mosaic images taken at multiple exposures (shutter) settings to provide high contrast inspection images, even in the intense sunlight. Joe Bibby and Rob Ambrose (JSC) played several videos of the Centaur system conducting field geology, handling samples, and manipulating softgoods while driving over modest terrain. Chris Culbert gave an overview of the approach taken by the team, and links to the LAT and other ETDP projects.

After answering a few questions, the weather cleared enough for the team to reset their experiment, and give it a try. For the rest of the day, small showers erupted, and the team adeptly stopped and re-started their operations. The many days of test and practice paid off, as each test was completely successful.

The Coordinated Field Test used the SCOUT, Centaur, ATHLETE and K-10 robots as part of an EVA team, combining them with two suited subjects. The human-machine EVA team worked through a scenario derived from ESAS Sortie and Outpost requirements. The humans rode into the test site aboard SCOUT, driving it in an onboard control mode. Dismounting, they walked to the Pressurized Robot Container (PRC) mounted on an ATHLETE. The ATHLETE system was "kneeling", in a low stance with a small ramp that would allow the suited subjects to climb into the PRC, despite the 200lb packs they were carrying in Earth's 1g gravity. Once inside, the suited subjects began a recharge of their suits, as they would do in a emergency, or as a technique for extending their suit's range beyond the mass they were carrying on their backs. The PRC was outfitted with two donning stations for the suits, allowing them to recharge while the structure held the suit's weight, giving the test subjects a well deserved rest. During the human's recharge, Centaur and K-10 went into action, servicing the SCOUT rover. Centaur approached SCOUT, navigating through an obstacle strewn work site, and autonomously positioning itself at the rear equipment tray on SCOUT, using laser and vision sensing. Finding an EVA sample box, it reached, felt, grasped and extracted the box in a fully autonomous mode, turning to place the box on its rear work bench, and then driving clear of Scout's next move. The K-10 rover performed a 360 circumnavigation of SCOUT, taking images at multiple exposures to be used in vehicle inspection and EVA closeout documentation. K-10's Work was also autonomous, and it was able to avoid obstacles in its path. SCOUT was then driven to a new position in an un-crewed control mode, using GPS way points previously recorded and in SCOUT memory, driving out of camp and on to its next task.

This scenario demonstrated EVA roles for humans and machines, experimenting with four very different classes of robots. While these exact robots and suits will not go to the moon, the functions of driving, carrying, manipulating, and inspecting are fundamental. The command & control systems behind the scenes, while hard to see, are equally important. These technologies and approaches worked, but there were numerous things that did not work. Fortunately, Spring and Summer testing had found these problems, and the Engineers had found alternative solutions- but not before learning something new.

Following this successful run, the group broke for lunch and then returned for an afternoon of additional demonstrations. Guests were trained on the SCOUT onboard control, and given the chance to drive the rover over the desert terrain.

Centaur was driven back up to the hill top, and between rain showers, was setup for the ATHLETE rappelling task. During this work, and with multiple short-duration rain showers, the team innovated a new sheltering technique, with Centaur parking under ATHLETE with its tarp. During a dry spell, the entire rappelling sequence was executed, taking about 25 minutes. This was an improvement over the prior day's 40 minute run, and used many of the improvements identified by the team. Once Centaur had extracted the tether and was clear, ATHLETE selected and drove its screw anchor into the hilltop. Centaur then attached the EVA tether hook, using some finesse at the end to straighten the line, and then ATHLETE started its rappel over the edge. The demonstration was ended there, setting the stage for the FY07 work that will include further work in ATHETE rappelling techniques for exploring lunar craters.

The Engineers whose work had made this field exercise so successful were dismissed, later gathering at a local restaurant for a well deserved celebration. A casual observer would not have been able to tell they worked at NASA field centers separated by thousands of miles, having become a single team and a group of good friends.

This was the worst weather day of the field exercise, but the camp gear and shelter performed well. As many as 6 rain showers came across the site during the day. Approximately 0.5 inches of rain fell at the site in the mid-morning shower, with wind gusts of ~25 MPH. Weather came in from the West, and moved East all day. At night fall, an even larger rain storm caused flooding in Flagstaff and left standing water at the field site.

The next morning crews began disassembly of the test site and started packing equipment for the trip home.