Generic Oceanographic Array Technology Systems

BP_02 -- Masai_02
Sunday, May 26, 2002
Unpacking on Alliance. Assembly of Xanthos and Caribou initiated. Xanthos tested positively for all functions. Set up MOOS computer and temporary network.
Monday, May 27, 2002

Port all day due to bad weather in Framura. Caribou assembly completed.
Set up operation center in Alliance aft pier with the help of the ship crew who made penetrators for antenna cables etc. This is the perfect setup, allowing the operations team to have full visual control of the operations area. We also set up shop in the Programmer's Room, and Galetti connected our network on the bridge and in the programmer's room. Absoliutely ideal setup.
Deployed on fishing line for initial dynamic control tests inside Pagliari dock. Severe control problems identified, for both yaw and pitch. When Xanthos was closed up it turned out the aft sphere did not hold a vacuum. A lot of effort did not reveil the reason, but there was no way to use that sphere. Too bad, because Xanthos was otherwise ready to go. We did not have a spare aft sphere, but Jim and Rob suggested we try to modify the spare front sphere , which would require significant rewiring of the penetrators. It was also decided to ask the group arriving Saturday to bring a spare aft sphere in case the rebuild does not work, which is unlikely, because it involves splicing the radio and GPS antenna connections.

Tuesday, May 28, 2002
Port all morning due to bad weather.
1200: Transit to Framura in expectation of better weather. Slightly rough on arrival, and SACLANT decide not to deploy immediately. With vehicles untrimmed we decide not to deploy either until the weather improves.
Performed dynamic control runs in late afternoon, without success in solving pitch oscillations. Due to bad weather forecast it was decided not to deploy LBL transponders.
Wednesday, May 29, 2002

Bad weather in the morning. SACLANT deployed the OEX after lunch for sidescan surveys and we did not get Caribou into the water until after dinner.
1600: Deployed LBL network with 5 transponders.
1800: Got a few runs in after dinner. The vehicle operation from the bridge, and running MOOS through the radia link is excellent and efficient. However, we do not yet have the dynamic control problem solved.

Thursday, May 30, 2002

SACLANTCEN operated in the morning making sidescan surveys.
0100: Deployed Caribou and completed 66 missions throughout the day without solving the pitch control problem.

Friday, May 31, 2002

Beautiful weather all day. SACLANT operated in the morning, while we took over in the afternoon. Made 33 runs, and succeeded in keeping Caribou under water, but still severely oscillating in pitch. A systematic dataset was collected, though which should allow us to resolve the control problem through thorough analysis.
1930: Departing for La Spezia. End of BP-02.

Saturday, June 1, 2002

In port in La Spezia, working on the dynamic control problem in the dock. Some progress made, but still problems with pitch control.

Sunday, June 2, 2002

Four new team members arrive, Mike Bosse, Samm Desset, Rick Rikosky, and Chris Bohner with the spare aft sphere for Xanthos. . Joe Edwards called from the beach and said he would show up Monday. That was actually good because we only had two cars to drive 12 people up to Garfagnana for or BBQ. They had all taken the train from Pisa instead of a rental car, so the other passengers had looked at them a little strangely! 5 americans with
In the afternoon we had a great BBQ in Pieve San Lorenzo. Nice to get everybody off the ship for a day off and away from AUVs and computers.

Monday, June 3, 2002

Port all day, loading Center stuff and personnel. Too bad, another day with beautifull weather and calm seas.
Ran Caribou in Dock for dynamic control. Ran a series of missions, after which the data was uploaded via the wireless Ethernet with Caribou hanging from the crane, and analyzed. The analysis showed that the unstable pitch control was due to noisy depth sensor data, which contaminated the pitch-rate control. Paul had overnight implemented a depth-sensor filter, which eliminated the noise sufficiently to up the rate gain Kd by an order of magnitude.
1700: Requested one more mission from the crew, and the vehicle made its first close to perfect level dive without porpoising. Made one more with identical results. All we need to do now is have her dive a little faster, and we should be there! Also yaw control should be sharpened up, based on the analysis of missions in the dock.

Tuesday, June 4, 2002

Start Masai02.
Left port 0900
Slow transit, waited for Navy seals to come alongside.
Arrived at Framura 1130. Nice weather. Kinematic GPS not ready. Postponed missions to after lunch. Decided not to deploy LBL due to bad weather forecast.
Tested new dynamic control. Caribou showed nice level flight at increasing depth. Petre had added a couple of pounds of weight. Performed 2 200x200m surveys in cable area.

1800: She came slowly to the surface all day, but at completion of the last survey at 1800 she stopped, and then slowly sank, in 23 m water depth. All acoustic systems continue to work. Matt tracked them. Crew was great. Captain called Centre to get permission to have divers from Levanto work from Alliance workboat.

1900 Divers arrive on Alliance. Pete and Per Arne went out in the workboat with the divers and triangulated using the LXT and GPS form the bridge. The workboat was directed to the position.

1950: divers went down.

0805 Caribou was located 10 m from the spot marked by the bridge, secured and lifted to the surface. No flooding observed. Hypothesis was that she was trimmed close to neutral and hit some warmer water.

Wednesday, June 5, 2002
0800: Left Framura in the morning due to rough weather and arrived in La Spezia.
Caribou reconfiguration. SAS payload assembled and integrated in Caribou. Array fitted in nose.
1300: Prepared Xanthos for trimming and balancing.
1400: Xanthos trimmed off the fantail. All system checks ok.
1500-1630: Xanthos operated with fishing line from workboat in the dock. Fishing line got entangled at turn, and it was decided not to run her under those conditions. Instead she was towed, and Matt Grund demonstrated 2-way acoustic communication.
1600: According to Guerrini, the forecast for 5-terre is very bad until Saturday. Discussed the situation with the Captain and Lian Wang, and we decided to seek clearance for operation in Biodola Bay, Elba. Clearance for Thursday could not be achieved.
2000 Caribou assembled in SAS configuration, ready for trim in the morning. Don Eicksted kept working on DSP software to eliminate crash after 10000 pings. Believed solved at end of evening. Still needs to finalize MOOS interface with Paul. Expect to be in port all day Thursday as well, so plenty of time for that.
Thursday, June 6, 2002

0800: Weather beautiful in La Spezia harbor. Reports from Levanto say calming seas. Decision about whether to go to 5-terre will be taken at 1000. Decide to take the opportunity of trimming Caribou with new payload.
930: Caribou trimmed off fantail.
1030: Back on deck, preparing for transit to Framura.
1100: Partng for Framura.
1300: Arriving in Framura in 50 knot wind and high sea state. Decide to steam around until clearance for Elba secured.
1600: Clearance for Elba granted from June 7-17, working from 0800-1900. Steaming at low speed to Elba.

After leaving La Spezia Paul Newman, Don Eickstedt and Mike Bosse worked on the integration of the SAS system into MOOS and made amazing progress with Caribou tied down on the deck, and ran SAS missions with pinger and acquisition running. Set up data structure and SAS menu for MOOS. Joe Edwards analyzed data collected, and the system seems to work like a dream. Cannot wait to get Caribou back in the water in Biodola and collect real data!
2000: Mike Bosse and Done Eickstedt working on perfecting the SAS-MOOS interface.
Rough seas. John Leonard was hard hit by seasickness all day.
In the evening Mike Bosse and Don Eickstedt worked on the integration of the SAS into MOOS and found a few bugs. All day we had saltwater running over the SAS payload for cooling, but in the evening one of the crew turned it off without notifying us, and even though the temperature is monitored, nobody payd attention to it since it had beeen stable all day. Suddenly after all bugs were fixed at midnight, the DSP ceased to work! It turned out the temperature was 47 degrees in the payload. We got the pump back on, but too late. Could not restart the system Decided to turn it off and hope that it only needs to cool over night.

Friday, June 7, 2002

John Leonard had a tough night. We steamed slowly to Elba in sea state 10. A couple of others got seasick too.

0800 Start of operations in very nice weather in Biodola.
Deployed 4 node transponder network with GIB bouys on top, a communication bouy and 4 radar reflectors as SAS targets.
Prepared Caribou for operations. The SAS system started up ok after having stalled late last night after one of the crew shut off the saltwater pump and the payload overheated.
1230 Caribou launched for dynamic control missions. Had difficulty diving and Pete added some weight after obviously being more conservative after the Tuesday sinking incident. That helped, and she dove nicely, but at the end of the day she still had some pitch oscillations of about a meter.
We had the SAS payload section running on two set-point missions and several trackline missions, including one close to the 4 radar reflectors we deployed. All missions were very well executed apart from the pitch oscillations which we have to eliminate. However, a very good day, collecting the first SAS data with Caribou.

The GIB bouys did not interfere with any of our other acoustics systems, and the tracking, displayed on the OPS bridge, was spectaclar, and extremely useful for operations.

In the evening we had another human error. While Don was uploading data from the SAS, Caribou was turned off to be taken into the garage. It all happened because our usual 2000 science meeting was interrupted by the workboat coming back in with our transponders, so a couple of our people were not aware that the data upload would happen. Of course this was the one of a thousand crashes where un-repairable disk damage occurred! We decided to take Caribou apart in the morning and rebuild the system, while running Xanthos for navigation and communication work.

Saturday, June 8, 2002

Another nice morning in Biodola, after having steamed all night. We started up deploying the LBL net again, this time with a smaller, 250 m baseline, and the WHOI communication bouy. We did not deploy the targets since we did not expect Caribou back in the water today, due to the SAS repair.

Xanthos was run all morning doing LBL and communication missions. Matt Grund and Chris Bohner tested several new communication protocols between the ship, the communication bouy, and the AUV, including interactive requests for specific mission parameters to be transmitted back to the ship. As also experienced by the REMUS team last month, the propagation conditions are very harsh for acoustic communication, with a very strong surface channel in the upper 5 meter. In the afternoon the sound speed profile is close to iso-velocity, and the acoustic communication improves significantly.

At 1220 Xanthos was recovered to allow everybody to eat and watch Italy play in the World Cup. The lost 2-1 to Croatia, so the mood of the crew was not too good the rest of the day!!

After lunch the Xanthos D-cell battery packs were rebuild. She was ready for launch at 1630, and Paul Newman executed several picture perfect missions in the LBL net.. Following the last one we had Xanthos drive back to the ship on the surface while the workboat recobvered all navigation bouys in 30 minutes. When they came in with the last one Paul drove the AUV slowly alongside the ship 10 meter out, witing for the workboat to pick her up. The first mate Roland called up the Captain, and they were both duly impressed! It really was a great show!

In the evening the SAS software was being rebuild so the system can be back in the water in the morning. Unfortunately one of the penetrators has leaked and a pin is corroded away, but Don brought enough spares, and Pete Haqll is rerplacing it while Don works with the software.
After recovering the data from the SAS missions Friday, Joe Edwards and Rick Rikoski went to work on the analysis, and in the late eveining Rick found what we believe was the radar reflector targets we deployed. Tomorrow we will have Joe Edwards run his SAS imaging software on that data. This is really everything coming together, AUV control, navigation and acoustic sensing and processing. It will be really astounding, and due to the incredible team spirit if we pull this off the first day we had the SAS system in the water, in particular after we lost more than half of our time so far to bad weather!

Sunday, June 9, 2002
We decided to get an early start because the Centre wanted to run their AUV in the afternoon for a sidescan survey of Biodola Bay, and we all got together at 0700, except Don Eikstedt who had gone to bed at 0655 after having worked all night of rebuilding the software on the SAS payload! Joe Edwards had been up at 0500 and had helped Don put the system together and test it, and it was all ready to put back in the vehicle. Pete Hall did this while the Matt Grund and Chris Bohner did their daily 'milk run' and deployed the LBL beacons, the GIB pingers and the acoustic communication bouy. They are getting really sharp at that now, completing it all in less than one hour.

At 1000 Caribou was assembled and ready to go in the water when the wind picked up, and the Alliance was sliding on the anchor. Witth one workboat out of commission due to a broken gear box, and the other flaky, and stalling a couple of days ago so the boys had to paddle back (see picture June 7), it was understandable that the Captain would not let us deploy. Unfortunately the wind did not die down until lunch. The Centre were scheduled to deploy after lunch, but we should be able to run simultaneously. However, our LBL net interferes with their tracking system, so we had to pull our LBL transponders which Chris did in 10 minutes. We still had some work to do on the dynamic control, so no big disaster, the important thing was to keep the GIB tracking in. It has proved absolutely invaluable for our operations, providing high resolution navigation information to the Operations Center (OPS) on the aft bridge.

WE got the workboat at 1500 and deployed and made several trackline missions. Paul had set up a file transfer protocol so that he could efficiently send navigation and control parameters back to the bridge between missions, and that proved extremely effective, and after a couple of hours, Caribou was well trimmed in both yaw and pitch. We then ran the tracline with the SAS sonar on, and it worked like a dream. The previous days the Captain wanted the vehicles to be out of the water at 1800 in order to recover the moorings before we had to leave the area before 1900 when our clearance expires. However, today he came back to the aft bridge and said: "THAT vehicle you can keep in the water until 1845". Again Paul piloted the vehicle back to the shoip, first with a trackline mission with the sonar on, to within 100 m from the ship, and then under manual radio control alongside the ship until she was right under the crane! Most of us missed it because we were at dinner, but it was reportedly a great show, even though Pete as operations manager did not hesitate to express his concern regarding the risk of going THAT close to the ship!

At 1900 we left for initially La Spezia, but during the transit it turned out the equipment the Centre was to pick up was not ready, and we decided to go directly to Framura and start operations at 0800, with the Centre deploying to image the target field with their side scan vehicle. MIT was then to take over at 1100 deploying beacons etc, followed by AUV operations.

Monday, June 10, 2002
Arrived in Framura at 0800 with some swell, but otherwise beautiful weather and a light breeze. The Center started operations while we prepared our moorings for Matt and Chris' daily deployment run! At 1030 they were out of the water, and our moorings were deployed and surveyed at 1115. We decided to postpone our missions to after lunch, and instead had the workboat pick up John Leonard, who had spent a couple of days on shore to recover after the stormy transit to Elba.

After lunch Pete tested the emergency trackpoint transducer in Caribou and could not get it to respond. Also the strobe light was out. Both are safety measures you can run without, but obviously at increased risk if something goes wrong, And we decided to crack her open and check it out. It turned out the the trackpoint problem was actually the deck unit, and we borrowed the Centre's . The strobe could not be fixed and we borrowed the one from Xanthos and will have to switch it between the two since we have no spares.

Finally at 1500 we got her in the water and made a trackline test run in the LBL net. We had a scare first because the DVL did npot seem to work! Turned out it was because Caribou drifting into very shallow water, and no features on the bottom to lock onto. Once we siled her out in deeper water it was ok, and the mission was successful, and the analysis of the downloaded data suggested the the LBL was close to perfect.

Then we ran a survey mission in area B, where the Centre has identified a large number of concrete blocks protecting a cable. However, the DSAS sonar code would not start! Paul was "not impressed", and tried to debug via the radio link, but to no avail, and we decided to check out the sonar tonight on deck and instead run the surve using only DVL and GPS navigation, which was completed succesflly.

Then we tested a new behavior for launching missions via the acoustic modem from the ship, which was successfully achieved after a couple of attempts, with intermediate debugging via the radio. The operational setup we have developed now is extremely efficient for handling these situations, and Paul and Mike Bosse are a great pilot/co-pilot team. We were very concerned when Jay Dryer left, whose Navy background was crucial to the development of the piloting procedures, but Mike has really stepped up to the plate.

The final mission of the day was a survey in area A with the buried targets. We executed it with all systems on, running full closed-loop navigation with LBL,GPS, DVL and MRU, and acoustic telemetry back to the ship, and it was close to perfect. The mission aborted early at 1755 because we came too close to the bottom due to a noisy depth sensor when we apply the modem, but MOOS did the job!! We did our now perfected recovery "show" and 1820 everybody was at dinner! We still have a week to go and we are in the unusual position that both our vehicles are in top shape and ready to collect data! Now we just have to fix the sonar glitch tonight.

At 1900 we left for La Spezia to offload the Centre's vehicle so they can integrate their new SAS sonar which we need as well for insonifying the buried targets while our vehicle operates as a bi-static platform. We will then go back to Framura in the morning to resume operations around 1000.

Don Eickstedt and Paul Newman sat down after dinner to figure out what the problem was with the sonar. It was very weird , because the same code had worked perfectly yesterday in Biodola. Very quickly they found out it was a networking problem due to some changes made to vehicle computer, not the sonar, last night! Don was vindicated, his stuff had been ok all the time!

Rob Damus and Joe are working on integrating the autonomous detection algotrithms and the adaptive behavior changes into MOOS. We hope we will be able to test them one of the next couple of days.

All in all another great day, and we cannot wait to get back in the water again tomorrow!

  back to GOATS main page