Following the Ocean Infinity Search for MH370

The Background

In January 2018 it was announced that the Ocean Infinity company had reached agreement with the Malaysian government to restart the search for the wreckage of the Malaysian Airlines flight MH370 lost in March 2014. A description of the Ocean Infinity search equipment is provided here.

This page contains my observations on the progress of the Ocean Infinity activities in the Indian Ocean, based on the AIS data (ship navigation information) transmitted by the ship and collected by satellite. I will be updating the observations as the search progresses. The most recent comments will be at the top of the page.

I previously followed the 2014 to 2017 search by ATSB and Fugro for the same aircraft.

I am also tweeting on the search progress.

Richard Cole - Guildford, United Kingdom, January 2018


10th Feb 2018: The entry on AUV Sortie#3 and the early recovery of one of the AUVs on 29th Jan has been updated

The Search progress to date

19th May 2018: Seabed Constructor's search activities

Since Seabed Constructor started working in the Broken Ridge area some weeks ago I had not been able to discern the AUV search pattern being used. However, in the last few days a simpler search system seems to be in place, with the ship moving on the edge of the search area outside the 7th arc and the AUVs working back across the full width of the search area. The width of the search area was reduced to 81km some weeks ago, and the centre of the search moved 6km inside the line that was previously used.

An animation of the search in the period May 14th to 19th 2018 is shown below. The AUV track model uses the same AUV speed as before (6.6km per hour) but the gaps between the NW/SE AUV tracks is now 2.2km, rather than 1.8km used in earlier search activities. This may reflect analysis by Ocean Infinity that the sonar data is good to a range of +/-1.1km from AUV track or, possibly, the AUV is being operated further above the seafloor. In either case, the effect is that 20% more ground is covered per day. The position of one interpretation of the 7th arc is shown in the amended frame from the animation (but is not shown in the animation itself).

A feature has been added to the animation that shows the cumulative area of the seafloor covered as the AUV deployments progress.

27th March 2018: Animations of Seabed Constructor's activities

Below are two animations of Seabed Constructor's activities during its second swing in the MH370 search area. They are based on the S-AIS data transmitted by the ship plus my interpretation of the launch, track and recovery of the AUV units deployed by the ship. The AUV units do not transmit S-AIS position information. The animations indicate how the AUVs are used in the search.

The first animation covers the work inside the 7th arc, from the south of the secondary search area to the far north of the tertiary search area. There were a number of additional AUV missions deployed in the northern part of the search area (closer to Broken Ridge) beyond those shown but as their tracks could not be estimated accurately I have not included them. Note that the last AUV mission shown (W#28 in my numbering scheme) was followed by Seabed Constructor through the undersea canyon. Its exact path is unknown, presumably it moved left and right of the ship’s track – I have shown a constant ship to AUV distance.

The second depicts the work outside the 7th arc that followed immediately. Again there were a number of additional AUV missions deployed this time in the southern part of the search area but as their tracks could not be estimated I have not included them.

The ship moves off to the south during the 12th and 13th March, presumably to recover the ROV unit as reported by the Malaysian agencies.

The intention of the animations is to give an impression of the dedicated work being carried out by the crew of Seabed Constructor, from the point of view of a complete spectator.

18th March 2018, 07GMT: Working outside the 7th Arc to the end of Constructor's 2nd Swing in the search area

Seabed Constructor has finished eight days work in secondary search area outside the 7th arc. Around 17 AUV missions were launched in that period. The south of the area searched has waters deeper than could be searched with the equipment available to the earlier Fugro ships. A number of AUV missions searched this deeper area, though the search pattern used could not be easily discerned. On 12th March Seabed Constructor moved off to the south of the search area, leaving a number of AUV missions in operation. It was stationary was around 12 hours, presuably deploying an ROV (remote operated vehicle) dropped by cable to the seafloor. It was reported that another ROV had been recovered following an accident earlier in Seabed Constructor's programme in the search area and so the activity on 12th March was that activity.

The first map below shows the context of the search activity and Constructor's work in the south on 12th March. There are two detailed maps of the work over the period up to the ship's departure for Fremantle on 18th March.

10th March 2018, 12GMT: Completion of the Tertiary area search inside the 7th arc

Early on the 10th March, Seabed Constructor moved off to the southwest, presumably after completing the search activities in the north of the tertiary area.

In the last few days, Seabed Contructor had continued to work in the north of the Tertiary Search area on the edge of Broken Ridge undersea mountain range. The regular lateral scan has not been maintained (as far as I can detect) in the far northwest sector, probably due to the contours of the seabed, so it has been harder to keep track of what area has been covered. The movements of the ship indicate that a number of AUVs have been active in that northwest sector, outside the 7th arc. In this phase 28 AUV missions have been detected, but there may be several more in the that northwest sector. Ocean Infinity are getting quicker in the launch and recovery procedures it seems (particularly the latter), so that it can be completed between the available AIS data points (which can be an hour apart, dictated by the satellite pass schedule). I am not suggesting this is deliberate, just that it means some activities are not visible via the AIS data record.

In the north sector, just outside the 7th arc, a different search pattern has been used, parallel to the 7th arc. Three AUV missions have been been deployed covering the area between the north edge of the previously scanned area (scanned by Go Phoenix) and the north of the tertiary seach area. At the moment there appears to be an uncovered area just outside the 7th arc with no evidence of AUV movements in that area (that is launches and recoveries, or checks by the ship). This area was probably covered by the two missions indicated by blue arrows in fig 2 below, but the path of those missions was not clear. As a reminder, the AUVs do not make any transmissions that are recorded in the AIS record, their paths can be inferred from the ship movements if a regular pattern is being followed, which has normally been the case.

Despite poor weather, a final mission (#28) was launched through the undersea canyon with Seabed Constructor following at low speed (thanks to Don Thompson for pointing out this launch). Constructor broke off from following mission #28 to pickup the last two missions (numbers #25 and #27) in the water. Mission #28 was recovered on the morning of 10th March and the ship immediately moved off south-west.

The maps below give some views of this part of the search area.

Figure 2: Area just inside the 7th arc with the Fugro bathymetry survey shown as a background. Three AUV missions were lauched scanning parallel to the 7th arc. Mission #28 was launched last and was followed by Constructor through the canyon. Presumably, the AUV moved left and right of the ship track to check areas that could not be covered by the three other missions.

Figure 1: General view of the north of the tertiary area, inside the arc. The outline of figure 2 is shown in blue. The AUV tracks are indicated by broken green lines.The AUV tracks in the far northwest could not be inferred from the ship AIS data so are not shown.

6th March 2018, 21GMT: Status of search in the Secondary and Tertiary areas

In the last week Seabed Constructor has continued to move north through the secondary and tertiary search areas, working inside the 7th arc.

AUV search activities have continued using the rolling programme of keeping a number of AUVs in operation at all times, recovering and redeploying the units further north.

The search has now reached the far north of the tertiary area, where the undersea Broken Ridge starts to affect how the search can be carried out. Broken Ridge contains a number of volcanoes and steep ridges, which don't necessarily allow the simple linear scanning pattern used furher south.

A small update to the previous depictions of the lateral search pattern being used by OI: the AUV paths have been parallel all the way though this phase, rather than radiating from the centre of the 7th arc. The start points follow the curve of the 7th arc. This is a small difference but enough that the mid-mission checks performed by Constructor on the AUVs were beginning to not match the expected AUV paths, in the north of the search area.

The width of the AUV pattern has been widened so the AUVs are now operating from the 7th arc to the inside edge of the +/-25nm wide search area. As the seach is now further north than any of the work carried out by Fugro and Go Marine, search activities are closer to the 7th arc than at any time in the last three years.

Presumably, the search will continue around the north of the tertiary area and then south-west, outside the 7th arc.

The three detailed maps below show activities in the secondary and tertiary areas, working from south to north. The third map shows the contours of the northern section of the search area, the ridges and volcanoes of Broken Ridge. The positions of the AUVs are harder to predict in this situation as the AUVs may not be keeping to the simple search matrix as they manouvre to keep ideal sonar scanning conditions on the slopes of the sea-floor.

So far 20 AUV missions (each 26h long) have been completed in this phase and at least four AUV missions are currently underway. Since Seabed Constructor has at least seven AUVs ready for deployment (based on recent activities) it is surprising that more are not currently in use, which probably means I have missed some launches. There is no obvious indication yet in the pattern that AUVs have been deployed outside the 7th arc (i.e. to the east).

25th Feb 2018, 15GMT: Completion of Sortie#5 and change of tactics for the Secondary Search Area


Recovery of the six AUVs at the completion of Sortie#5 was delayed because of the bad weather. Seabed Constructor did not shuttle round the recovery area picking up the AUVs at the termination of their tracks. It would seem that the AUVs concentrate in one or a few areas to be picked up when Constructor is available to do so. Again, this may be improved tactics to deal with poor weather.

Following that, Constructor immediately started work in the Secondary Search area. The area to be scanned here is much wider, 40km (figure 1). The AUVs tracks are now at right-angles the 7th arc, rather than parallel, and individual AUVs are each assigned a roughly rectangular area to check (figure 2). The AUVs are being picked up, serviced and redeployed further to the northeast in a continuous process, moving the search in that direction. The AUV  deployments are on the eastern edge of the search area and Constructor continues the practice of checking each AUV roughly halfway through its mission (which will be 2.5 days long, if my track interpretation is correct). The complete expected mission of AUV#2 (for example) is shown in figure 3.

This search pattern is similar to the one described in an Ocean Infinity press release some months ago (figure 4 below).

Ther is a smaller amount of work in the far south of the area, where it overlaps with the previous Fugro search. One or more AUVs appear to have been deployed in this area (on tracks parallel to the 7th arc), perhaps to fill-in areas missed before.

Barry Carslon contributed to this analysis.

Figure 1: Summary of the status of the search. The tracks of the six AUVs deployed in Sortie#5 are indicated in green. The lateral search pattern in the northern Secondary Search area is indicated (rather faintly), better seen in figure 2.

Figure 2: Summary of the work in the Secondary Search area in the last three days. Six AUV deployments are indicated in the northern part of the area and at least four in the southern part (this covers more that one deployment of several AUVs). The AUV positions are estimated by a similar process to that used in the earlier sorties, based on a hypothetical AUV pattern and known AUV speed, confirmed by the checkpoints performed by Constructor during each deployment. For example, at the moment AUV#1 has two more lateral lines to scan before it reaches the first line scanned by AUV#2, which will take 12 hours. On that basis, recovery would be on the west side of the search area. Note that the models for the predicted positions for AUVs #5 and #6 have yet to be validated by checkpoint visits from Constructor. An expanded map of the search area from the Malaysian reports is shown as background. 

Figure 3: The search area with the expected AUV#2 mission shown the the dark green line. Slightly later data that shown in figure 2.

Figure 4: Ocean Infinity image of their mission planning software tool. The current search pattern is similar to the activity displayed.

21st Feb 2018, 07GMT: Launch of AUV sortie#5


After the clearance of the weather and completion of some unspecified activity, Constructor moved back inside the 7th arc and launched  six AUVs to the northeast. This would appear to complete the search of the primary search area within +/-25nm of the 7th arc, the originally stated goal. The protocols to launch the AUVs seem to have been shortened, presumably to save time and also allow work in poorer weather. The sea-state became poor again immediately after the launch of the AUVs but this did not stop the mission proceeding, since the AUVs are working many kilometres below the surface.

17th Feb 2018, 12GMT: Activity during the second swing in the search area


In the last 36 hours Constructor may have launched three AUV missions, inside a small part of the area it scanned during AUV sortie #2, on the first swing. The first launch shown doesn't follow all of the normal pattern for an AUV launch/mission start, so is indicated with a query. All the AUV missions lasted less than one day, rather than the normal two, if they were recovered (likely but not certain) before the bad weather arrived.

It also made another visit to a point further north where it has been several times before. This was also the point that Constructor first stopped in the search area back in January, so must have been chosen before arrival.

Weather has now closed in, probably sea-states will remain too high for AUV launch till late Sunday 18th.

15th Feb 2018, 21GMT: Return to search area

Seabed Constructor arrives back in the search area and stops at a location not previously identified as being visited, probably due to the AIS data black-out at the end of its first swing. The coincidence of the return with the track out of the area at the end swing #1 suggests it has returned to exactly the spot it was working before departure on 4th Feb.

13th Feb 2018, 20GMT: Return to search area

Seabed Constructor left Fremantle on 11th February and is currently on a course to the southern end of the primary search area, which it left in some secrecy 10 days ago. However, its course could change at any time, so it is not definite that is the final destination. It might have been expected that the search was about to move north to the designated secondary area, since the reports indicated the primary area had been covered by the work of the last 'swing' (tour in the search area).

Earlier today Seabed Constructor was stopped for seven hours in water 5500m deep (the first deep water crossed since leaving Fremantle). It was in a station-keeping mode that holds the ship's position with respect to the seafloor to an accuracy of a few metres. In the past the search ships have used this mode for deployment of Remote Operated Vehicles (ROVs) that are lowered by crane to the seafloor.

10th Feb 2018, 07GMT: Search status

Seabed Constructor arrived in Fremantle early morning 8th Feb local time. No further technical information on the search has been released. Several press stories had suggested that Swire Seabed, operators of the ship, have been given permission by the Norwegian authorities to operate at sea with the AIS transmitter turned off, which would prevent position information being available. Pressure on the crew from press and social media was quoted as a safety concern. If this is fully followed through, then AIS will be on for the ship's transit to the search area but will be turned off for all the operations once there. Note: I have never attempted to contact Ocean Infinity on any matter.

Constructor is due to leave port on Sunday 11th Feb and arrive in the search area about four days later.

6th Feb 2018, 22GMT: Interpretation of Malaysian Weekly Report #2

The Malaysian Authorities today issued their weekly report which included images of two seafloor Points of Interest (POI) identified in the search to date. Both were identified as geological and therefore of no interest in the MH370 search. The positions of the two POIs are indicated by blue boxes in my map of the primary search area. The area searched to date by Ocean Infinity is shown by the broken lines. The ship did not visit either POI when its AIS position transmission was enabled (the ship came close to the eastern POI, but only at high speed), so must have visited them when AIS was disabled between 31st Jan and early on 4th Feb. It would seem that the ship will in future visit such POIs only with AIS disabled, presumably to avoid disclosing the position of the wreck of MH370 if were to be found.

The ship is due to arrive in Fremantle around 21GMT on 7th Feb.

4th Feb 2018, 22GMT: Constuctor AIS data re-enabled

Early on 4th Feb (GMT) AIS data transmissions from Seabed Constructor were re-enabled. The ship is on its way back to Perth. A back-projection of its course suggests it had left the southern part of the search area where it first started work, but this is speculative. The green mark indicates Constructor's position when it disabled its AIS data late on 31st Jan.

At current speed, Constructor will be in the vicinity of Perth at 12GMT on Feb 7th.

3rd Feb 2018, 0900GMT: Position of shipwrecks found in previous Fugro search

There has been some speculation about Constructor's last reported position and those of the shipwrecks detected by Fugro in the earlier search. The first wreck was characterised by images of an anchor on the seabed, the later one by its extensive iron framework and identified tentatively by others as the S.V. Inca, lost in 1911. From my observations of the ships' movements at the time, both wrecks are located well to the south of the current search area.

The data blackout from Constructor continues, now 60 hours since the last reported position. It seems unlikely that more AIS information will be transmitted before the ship leaves for Perth, departure expected around 06GMT on 3rd Feb.

It seems possible that Ocean Infinity will decide to keep the AIS transmitter off for all future operations in the search area, while re-enabling it for transitions to and from Perth. There is so little ship traffic in the area that the decision could be defended on safety grounds, it seems to me.

1st Feb 2018, 1900GMT: AIS data blackout from Seabed Constructor continues

The data blackout continues, now 24 hours since the last position information transmission from Seabed Constructor. It seems clear that that the ship is checking out contacts (i.e. points of interest) in the areas they have scanned. Fugro needed to specially bring an AUV to the search area to perform these close-up checks so they were done in batches. Ocean Infinity have the required equipment to hand on Seabed Constructor so the checks are being performed as soon as the processed first-pass sonar information is available. Ocean Infinity also seem more sensitive to revealing the position of these contacts while Fugro/ATSB were open and allowed the search ships to keep their AIS transmitters enabled during the close-up checks, which in practice meant the precise locations were public.

Constructor has to leave the search area in about two days to meet its schedule arrival date at Fremantle, for crew change and reprovisioning. We will have to see if the position data transmission is re-enabled before it departs for port.

1st Feb 2018, 0700GMT: No AIS data from Seabed Constructor

The two services I subscribe to have not recorded any AIS data from Seabed Constructor in the last 12h. The services have recorded information from other ships in the same area, so it doesn't seem to be a general system issue. AIS services have improved in the last three years and the ORBCOMM system, used by several of the services, provides AIS data via a fleet of spacecraft about every 15 minutes. While ORBCOMM receivers can be saturated with AIS transmissions from hundreds of ships in very busy areas (as it was when Constructor was off the US coast) that does not apply here, and as noted other ships are being recorded regularly. The updated information recorded below includes the last transmission.

31st Jan 2018, 1900GMT: Rescanning part of the the area of Sortie#2

Rather than move into the green Secondary search area, Seabed Constructor has moved back into the Primary search area outside the 7th arc. It appears to be rescanning just one of the tracks previously scanned in Sortie#2. On this occasion Constructor is directly following the single AUV, so the AUV track will be very similar to the ship's and doesn't have to be predicted in the same way as described earlier. The ship's speed has been variable, suggesting that an AUV (if one was launched) has not been using its Side Scan Sonar continuously, since it is understood that requires a particular speed, around 3knots. The reason for the circular manoeuvre to the northeast is not understood.

The ship will be leaving for Perth in a few days, so there may not be time for a Sortie#4 with the full set of eight AUVs.

30th Jan 2018, 2030GMT: Update on Seabed Contructor and AUVs, end of Primary search area work inside the 7th arc

Constructor finished Sortie#3 in the primary search working inside the 7th arc. Only six AUV pick-up points were identified in the recovery area (figure 2 below). There was no recovery on AUV track number 7 as identified in the launch phase, at a time that would match to when AUV#7 was launched. It is possible AUV#7 was recovered early in its mission. The ship did stop briefly 12 hours after AUV#7's launch, close to the predicted location of AUV#7 at that time (see figure 1 below). After completing the (six?) recoveries Constructor began a move back inside the 7th arc, presumably to continue the search in that area (figure 3).

[updated 08GMT 10th Feb with information on the possible AUV#7 early recovery]

Figure 1: AUV#7 possible early recovery point (purple mark), 12 hours after launch

Figure 2: Recovery of six AUVs at the end of Sortie #3.

Figure 3: Movement of Seabed Constructor to new search position outside the 7th arc

29th Jan 2018, 00GMT: Update on Seabed Contructor and AUVs status, now working inside the 7th arc

As expected, Constructor moved to the inside of the 7th arc to search the remainder of the primary search area, as designated by the Malaysian Authorities. The ship has launched seven AUVs in the last 24 hours on what I have now termed as Sortie #3. From the extension of the Constructor's path to the northeast, AUV#3 seems to have been launched and then recovered, and then launched again on the same track. The time required for this repeated process caused an extension of the launch phase. After lauching AUV#7 Constructor then left at high-speed towards the recovery area at the northeast end of the primary search area.

To repeat, I am not tracking the AUVs via any transmissions they may be making. I am analysing the position, course and speed of the support ship to identify when the AUVs are being lauched. The pattern of the launch and recovery points, the primacy of the 7th arc as a datum, and the known operational speed of the AUVs allows a prediction of the AUVs subsequent location at any time. This method was successful in analysing the first two sorties of the AUVs.

27th Jan 2018, 1400GMT: Update on Seabed Contructor and AUVs status

It was seem that the AUVs have all been recovered. Constructor has moved north to a site (red circle) that was visited when the ship first entered the area. The position stability being maintained (a few metres) suggests the ship is lowering its Remotely Operated Vehicle (ROV), possibly to recover from the seabed what others have suggested to be a calibration target of some kind.

Since the target area has been covered, it is expected that the ship will subsequently move onto another part of the search area, perhaps inside the 7th arc.

26th Jan 2018, 2100GMT: Update on Seabed Contructor and AUVs status

The AUVs were redeployed from the 24-25th Jan (fig 1 beow) and are now moving southwest parallel with the last set of AUV tracks, outside the arc and also outside the primary search area as declared by the Malaysian authorities (fig 2). My analysis indicates that seven AUVs were deployed in this phase, rather than the eight on the first session. Possibly one was not ready for the deadline.

As of 18.30GMT on the 26th the first AUV units are being picked up at the south end of the search area (fig 3). This recovery phase is expected to continue until 11GMT on 27th Jan.

Fig 1: AUV redeployment at north end

Fig 2: Second set of AUV tracks from north to south, outside the yellow primary search area

Fig 3: Start of AUV recovery at the south end of the area. AUVs #1 to #3 recovered at 19GMT 26th Jan

24th Jan 2018, 08GMT: Update on Seabed Contructor and AUVs

The AUVs are now completing their missions at the north end of the primary search area. Constructor appears to be moving across the recovery zone picking up the AUV units as they arrive; the last is due at 15GMT (on the basis of the analysis described in a previous post below).

Maersk Mariner rendezvoused with Constructor around 00GMT and has now left the area.

23rd Jan 2018, 07GMT: Update on Seabed Contructor and AUVs

Following the idea of AUV position prediction in the previous post, Constructor has spent the last 12h moving NE through the AUV fleet, checking at least some of the units in passing. Maersk Mariner has turned NW, presumably to the meeting point.

22nd Jan 2018, 19GMT: An attempt to decode the AUV Search pattern

Seabed Constructor has been moving in the search area in a similar way to its activities off Durban a week or so ago. I attempt here to understand what is happening.

The ship has been following a repetitive sequence of steps:

1. It stops for a period,

2. then moves at roughly the AUV's speed (5kph) for about 1 hour,

3. stops again,

4. then moves off at high speed to repeat the process.

The direction of movement at AUV speed (step 2) does not seem to correspond to a consistent search pattern. The clue seems to be that as the AUV descends, its navigation system needs to be updated by the mothership so that the AUV knows exactly where it is when it starts scanning at depth. This process is described in a paper here. I am hypothesising  that this update process is terminated at step 3, after which the AUV moves off to carry out its mission. If a grid of arcs (parallel to the 7th arc) are drawn through the set of points that correspond to the set of step 3's, the result is eight lines each 1.6km apart. This path separation matches the known swath width of the sonar in the AUV.

On this basis, in the diagram the set of AUV release points (from the ship) are indicated plus the set of points the AUVs starts their missions (pink arrows). The times of these points allow a prediction of where the AUVs are at the current time on the basis they move off at the known AUV operational speed. These predicted points for 19:00GMT are indicated by the black circles in the maps.

The AUVs are moving northeast, covering the section of the primary search area outside the arc.

Seabed Constructor is now moving at moderate speed to northeast. It stopped a few minites ago as it passed the predicted position of AUV#7 at that time, which supports the theory.

21st Jan 2018, 22GMT: New vessel joining the search?

Maersk Mariner left Fremantle some days ago with the declared destination of Seabed Constructor (thanks to Barry Carlson for that information). Its projected course indicates it will reach the search area in less than 48hours. The ship's role in the search is not known at this time; there is a report it may just be taking supplies to Seabed Constructor. On departure from Fremantle, Mariner's next port was declared as Dampier in Western Australia, not back at Fremantle, suggesting no continuing role in the search.

21st Jan 2018, 12GMT: Seabed Constructor stopped in the search area

Seabed Constructor has stopped at the south end of the search area.

Sunset at that location is 13.15GMT.

19th Jan 2018: Journey to the MH370 search area

Seabed Constructor is moving almost due east towards the most southern point of the Southern Indian Ocean search area identified by the Australian Transportation Safety Board (ATSB) in late 2016. The ship has dropped back to a slightly lower speed in the last 24 hours and will reach the search area around 12GMT on Sunday 21st Jan, on the current speed and course.

The Malaysian authorities have published information on the areas that are discussed in the contract with Ocean Infinity. These areas are based on the ATSB designated 'First Principles' review in late 2016. The map below overlays the Malaysian information across the search status at the end of the 2014-2017 ATSB/Fugro/Phoenix International search. The map shows the currently projected track of Seabed Constructor as of 21GMT on 19th January. This indicates the ship is directed towards the southern segment of the 'Primary' search area. It is understood that Ocean Infinity will start with the Primary search area and progress to the secondary and tertiary ares, and then beyond. However, the details of what search pattern Ocean Infinity will finally execute remains to be see.

January 2018 - Deep water test south-east of Durban

Between Jan 6th and 13th 2018 Seabed Constructor carried out a deep water test of the eight Autonomous Undersea Vehicles (AUVs) that it will be using for the MH370 search. The notes and maps below describe the ship's movements during this period with my interpretation of what was happening - the division between phases is mine. Figure 1 shows the movements over the total period against the undersea contrours. All the work was at depths greater than 5000m.

Jan 6th  - 8th - AUV Deployment Phase

For 48 hours Constructor was moving in 6000m of water (figure 2). It is believed that a number of AUVs were released on a pattern with the pitch shown in figure 2. AUV releases would have been at points where Constructor was reporting stopped. A possible set of six release points are marked AUV-RP#1 to #6 in figfure 2, but this is speculative. 

Between pairs of release points the boat moved initially at AUV speed, presumably following the AUV, returning at higher speed to the next release point.

Following that, Constructor revisited (broken line in figure 2) the areas where the AUVs had been dropped, probably to check on their  status. It is believed that none of the AUVs were recovered at those times as their missions were not yet complete.

Figure 1: Overview of deep-sea test area

Figure 2: AUV deployment phase, Jan 6-8th

Jan 8th to 11th - Second phase, AUV Deployment in the north and AUV Recovery in south

In this phase the area of operations was increased (red line in figure 3) - the outline of figure 2 is shown as a broken line. Possibly, several more AUVs were deployed in the north, where the water is deeper. The ship also returned to the original area, presumably to pick up the AUVs released in the first phase at the end of their 48hr missions. The weather was poor in this period, with a large swell, and the ship probably had to delay some recoveries. As times the ship was moving down the swell at 1kn (1.8kph) for extended periods.

Jan 11th - 13th - Last Phase, purpose less clear

The ship moved off the east but returned west after moving 40km (green line in figure 3) - the purpose of this leg is not clear. It seems further AUV missions were then deployed and recovered. The legs to the south-east were probably a holding pattern while the missions were completed. The ship left the area moving to the east at 08GMT on the 13th, on a course that would take it to the MH370 search area.

Figure 3: Operations over 8th Jan to 13th Jan

6th Jan 2018 - Departure from Durban

Seabed Constructor docked at Durban between December 27th and Jan 2nd. On 6th Jan it completed a move to deep water south-east of Durban to test the eight AUVs it is carrying onboard.

December 2017 - Operations in mid-Atlantic

Between December 13-17th Seabed Constructor carried out operations in mid-Atantic, around the reported position of the wreck of the German blockade-runner Weserland, sunk by the US Navy in 1944.

The work seemed to be in three phases:

1. A Search phase of 50hr, starting at arrival

2. A Recovery phase of 21hr

3. A Stationary phase of 6hr, ending at departure for Durban

The ‘recovery’ phase has five points where the ship is stopped for two hours on each occasion. These points are symmetric with the historic wreck location and I hypothesize that five AUVs were recovered at these times. On that basis the area has been divided into five roughly equal areas, each of

The search phase of 50hr corresponds to the endurance of the Hugin AUV. If I assume that the total contiguous area was searched in that period then the achieved rate is AUV, for the time in water. The additional time taken to recover and refurbish the AUV is not included in this rate.

The points of deployment of the AUVs were not identified in the S-AIS data, perhaps because the ship is only stopped for a short time, while the recovery takes much longer (2hr).

In the ‘stationary’ third phase the ship’s position was stable to a few metres, which may indicate that the ROV is deployed. This behaviour was seen in the Fugro search phase ROV deployments.

The two diagrams indicate i) Phase 1 and ii) Phases 2 and 3. In the second diagram I have sketched an indicative AUV track which would take 43hrs to execute at the expected AUV speed and pitch.

The recorded detailed movements of the ship in the centre of the search area are not fully explained. The ship moves over the same area twice, whether tracking one or two AUVs, or doing something else, is not clear. A sixth AUV may have been deployed and recovered before the ‘recovery’ phase identified above but is not clear in the data.

The Search by Fugro

Between 2014 and early-2017  an extensive search of  an area in the Southern Indian Ocean was carried out by a number of parties but largely the Dutch Fugro company. The final state of the search is shown in the map below.  The map  also shows a priority further search area recommended by the Australian Transportation Safety Board  in December 2016  but not carried out.

I followed the search via the AIS position information transmitted by the search ships and produced a number of status reports.