After the accident to Air France 447, which crashed into the ocean off the coast of Brazil in 2009, the data in these handshakes was modified to allow for a rudimentary form of track reconstruction. In the simplest terms, the airplane and the satellite communicate using electronic signals. The speed of the signal transfer is known, and the position of the satellite is known. With that information, experts can calculate the airplane’s location based on the time it takes for the signals to travel back and forth. Fortunately, the pilot’s actions to cut off all electronic contact with MH370 failed to fully shut down the airplane’s capability to contact that system.
A dedicated group within the official investigation team, led by the Australian Defence Science and Technology (DST) Group, conducted an analysis of these handshakes. Using sophisticated and groundbreaking methods, they were able to reconstruct the basic path of the airplane as it flew to the southern Indian Ocean. The results of their work were used to direct the official investigation’s search for wreckage. Resources were directed to the areas of the ocean identified as most likely to contain the wreckage.
For those who are interested in scrutinizing the work of this group, they have produced a number of detailed reports that are available through the Australian Transport Safety Bureau website (www.atsb.gov.au). In those reports, they explain their logic, and procedures, and calculations. They also include the assumptions they used in producing their results.
Two assumptions that were used in their calculations were: 1) MH370 was flying as an unpiloted airplane, with no pilot input; and 2) the airplane eventually ran out of fuel at high altitude, and descended at high speed into the ocean. As we have seen in this account of what happened, there is proof that these two important assumptions were not correct.
There is an appearance of “certainty” (my word) in the DST group calculations. I use the word “certainty”, because with their assumption of an unpiloted airplane, they are able to make mathematical calculations, and mathematical calculations produce definitive results. The results can be accurate, provided they are calculated using accurate inputs.
By using their unpiloted airplane scenario, the DST group could assume that because the airplane was flying on autopilot, with no pilot inputs, it maintained a constant altitude and airspeed during the flight to the south. Based on an assumed power setting and altitude, they could calculate where the airplane would run out of fuel. Then, using a B777 simulator, they could establish how far the airplane would be able to travel after the fuel was exhausted. Finally, by factoring in the calculated airplane tracking information they had extracted from the handshakes, they could establish boundaries to define the specific search areas.
While the official investigation’s active search for wreckage was ongoing, the investigation authorities publicly presented a high degree of confidence in their assumptions and calculations. They expressed a high degree of “certainty” about where they were searching, and the likelihood of finding the wreckage.
The DST group’s unpiloted airplane theory is invalidated by the proof that there was a pilot controlling the airplane. There was a pilot-controlled descent, and a pilot-controlled ditching. Both of these events occurred before the airplane ran out of fuel. If you accept the evidence that a pilot was controlling a functional airplane, then you have to accept that the DST group calculations are not valid.
Unlike the autopilot, the pilot can (and did) take independent actions. Without the “certainty” provided by the unpiloted airplane assumptions, it is not possible to use the handshake data to predict where or when the airplane started to descend, or which direction it travelled during its descent, or its airspeed during the descent, or its rate of descent.
During the flight south, the pilot could have changed altitude, or changed the power setting. That would affect fuel consumption, and maximum range. It is not possible to know how much fuel remained in the tanks after the last handshake. Therefore, it is not possible to accurately predict how long, or how far, the airplane flew after the last handshake. It is not possible to establish the time the airplane entered the water.
With the uncertainty introduced by a pilot-controlled scenario, it is not possible to predict a touchdown location for the airplane, other than in very general terms encompassing a wide area.
There is no doubt that the DST group calculations prove that MH370 flew a track to the southern Indian Ocean. The drift modelling completed after the discovery of the wreckage pieces confirms that is where MH370 is. The basic track line established by the DST group most certainly points to the general area of the southern Indian Ocean where the wreckage of MH370 is located. Unfortunately, that basic track line, pointing to a general area, is the extent of what can be determined using the handshake data alone.
When (and I guess if) the official investigation accepts that MH370 was a pilot controlled event, perhaps the experts in the DST group, or in some other group, will use this (new to them) information to restart their calculations. They would have to come up with new assumptions about the flight south. They would have to account for extensive variables regarding altitudes, and airspeeds, and power settings. They would have to assume the pilot could have changed anything, at any time.
It is reasonable to assume that the pilot was aiming to ditch the airplane within a specific target area on the ocean surface; at a location where his pre-planning told him the underwater topography would be least conducive to the wreckage being discovered. Any new calculations would have to make new assumptions, based on a fully flyable airplane. Assumptions would have to be made about where the pilot started to descend from a cruise altitude, and about his rate of descent, and about his airspeed during the descent, and about where and when he made any turns.
The pilot wanted to ditch the airplane with the least possible damage. To prepare for the controlled ditching, he could have turned the airplane into the wind. He could have made a heading change to account for the waves and swells.
I will now address an aspect of the relationship between the investigation authorities and the DST group that I believe requires particular focus. There is at least the possibility that the investigation authorities knowingly influenced the work of the DST group to support a particular outcome. Was the official investigation looking to the DST group for extra ammunition to support their selection of the search zone? Were they looking for credibility to justify their ongoing search?
In their document titled MH370 – Search and debris examination update (dated 2 November 2016), the ATSB referenced conclusions from the DST group that were very helpful to their ongoing wreckage search. Specifically, the DST group (incorrectly) gave the official investigation hard data pointing to a high-speed dive. This evidence was convincing, because it appeared to come from independent experts, who were working with hard evidence. (Supposed) hard data now (unjustifiably) confirmed what the official investigation had been saying; that MH370 entered the water in a high-speed diving crash.
The official investigation framed it this way: that the final satellite communications to and from the aircraft showed a high and increasing rate of descent. The DST group’s hard evidence showed a rate of descent approaching twenty-five thousand feet per minute – basically, a classic spiral dive.
In that same 2 November 2016 update, where the official investigation released this hard data, they also released the results of their examination of the recovered flaperon, and the recovered section from the right outboard flap. The witness mark evidence they found on both the flaperon and the flap section (incorrectly) confirmed that both had been retracted (up) during a high-speed impact.
These two quite astounding claims: 1) that they had hard data showing a steep dive; and 2) that they had physical evidence of flaps retracted, were particularly influential in countering the views of those (including me) who were saying that MH370 was a pilot controlled event.
I knew at the time they released their update in November 2016 that the two claims they were making were clearly false. I remember wondering, as I still do, if everyone on the investigation team believed them to be true. Surely some members of the investigation team knew that what was being called hard data was in fact derived data. Hard data comes directly from something like a flight data recorder. All their results were calculated within significant margins of error, using assumptions and interpretations.
In any event, the release of this supposedly unimpeachable information from the ATSB was widely covered in the media, and generally accepted, and it supported the ATSB’s logic for their ongoing search for wreckage.
There will be more on this later, but first we can look at how the DST group came to support the inaccurate conclusion of a high-speed dive. Thankfully, we can do so without delving too deeply into the complex science they undertook. Here are the basics.
To keep our task manageable, we will not go through the complexities of challenging the accuracy of the DST group’s calculations that produced the first six handshake locations. Many people have already challenged the accuracy of the track line they came up with for the flight southbound. However, for simplicity, we will not challenge that. Instead, we will start from handshake #6. (Actually, I should say “the location of” handshake #6, but that makes for awkward writing, so I will drop the words “the location of”).
The time between handshake #6 and handshake #7 was only eight minutes. The official investigation contends that during those eight minutes, the airplane ran out fuel, and the engines stopped. To be more specific, they claim the engines stopped about six minutes after the airplane passed handshake #6. If you do the math, this is two minutes prior to the airplane reaching handshake #7.
The engines drive the electrical generators that supply electrical power to the airplane. With no electrical power, the communications unit that was producing the handshakes would shut down.
A quick reminder here: the B777 has an auxiliary power unit (APU), which is a small jet engine located at the tail end of the fuselage. In the air, it is normally not running. It is used only as a backup, but when it is in use it can supply power to (among other things) the airplane’s electrical and hydraulic systems.
The official investigation believes that when the engines stopped from fuel exhaustion, the APU automatically started. It is designed to do that. The APU uses the same fuel as the engines, but it has access to the last bit of fuel in the tanks – fuel that is not available to the engines. The official investigation contends that the start-up of the APU restored electrical power to the communications unit, and when the communications unit rebooted it sent a log-on request, and that is why handshake #7 happened.
To the official investigation, the reboot of the communications unit is proof that there was an automatic start-up of the APU, and the automatic start-up of the APU is proof that the engines had stopped, and the stopping of the engines is proof that the airplane had run out of fuel.
We know that none of that actually happened, but from a theoretical/technical perspective, it certainly is possible for it to happen. What seems to have made fuel-starvation even more believable to the official investigation is that the (supposed) start-up of the APU happened very close to the time that they had calculated for when the airplane would run out of fuel (based on their assumptions).
According to the DST group calculations of the handshake data, the unpiloted airplane was already in a descent before it reached handshake #7. Naturally, they attributed this descent to the loss of engine power. They assumed that by the time the airplane passed handshake #7, it was already out of control.
The next part of their calculation (to determine where the airplane crashed) was to insert the results of Boeing’s no-engines simulator testing. This allowed them to establish the parameters for their wreckage search area. To them, everything was logical, and everything fit together.
The problem is that all of their calculations were based on the unpiloted airplane theory. We know that the unpiloted airplane theory has been disproven by the evidence presented in this book. We know that the pilot was controlling what happened to the airplane, from start to finish. We know that the pilot was controlling the airplane all the way until it entered the water.
With no flight data recorders, it is not possible to determine the actual sequence that the pilot followed to prepare for the controlled ditching. However, it is certainly possible to come up with potential sequences that would account for the satellite data that was interpreted by the DST group as showing an uncontrolled high-speed dive. The following describes one such sequence.
Let us assume that the pilot used the autopilot during the flight south. This is a very reasonable assumption. Let us accept that the airplane followed the route calculated by the official investigation that took the airplane to handshake #6. As mentioned, I believe the accuracy of that track can be challenged, but let us accept it for now.
As he approached the end of the flight, the pilot would be keenly aware of the airplane’s fuel status. As part of his pre-planning for this event, on other flights he would have paid particular attention to watching fuel consumption numbers. He would know almost to the minute how much fuel (flying time) he had remaining. Given his intention to ditch the airplane with the least possible damage, he would have had absolutely no reason to allow the airplane to run out of fuel prior to reaching the water surface.
We can imagine that as the airplane was nearing the location for handshake #6, the pilot would be preparing for the controlled ditching. (I remind you that the pilot would have no concept of any handshakes – he thought he had shut off all contact with the airplane – I refer to the handshake locations here only to allow a comparison with the official investigation’s version of events.)
Whatever happened to precipitate handshake #7 (we will accept for now it was a reboot of the communications unit), it could not have been a result of the APU automatically starting after the engines stopped. We know the engines did not stop. Therefore, the reboot of the communications unit must have been a direct result of some action taken by the pilot.
From an overall investigation perspective, it is not necessary to figure out exactly what precipitated the reboot of the communications unit to come to a conclusion about the loss of MH370, but for general interest, we can speculate.
In preparing for the controlled ditching, perhaps the pilot was reconfiguring the airplane’s electrical system to restore some electrical function he had shut down. Or perhaps he was changing some electrical system to a different power source. In his reconfigurations, perhaps he somehow managed to interrupt, and then restore, electrical power to the communications unit.
In my view, it is more likely that the pilot started the APU intentionally, and then switched some electrical services to the APU. Perhaps this caused power to the communications unit to be interrupted and then restored, causing it to reboot.
There is logic for why the pilot might start the APU for a controlled ditching. Perhaps he wanted to make sure he would still have electrics and hydraulics if the fuel to the engines ran out at the last minute. Having the APU already running would protect against the risk of losing electrics and hydraulics during the controlled ditching. He would know that the plumbing inside the airplane’s fuel tanks would allow the APU to have access to the last bit of residual fuel in the tanks; fuel that the engines cannot access.
Another possibility is that the pilot started the APU as part of his plan for maintaining control during the aircraft’s entry into the water. He would know that in a controlled ditching, the engines would touch the water first, and they would stop immediately from water ingestion. He would know that the APU, which is higher up in the tail, would keep running after the engines stopped. The APU would maintain electrical and hydraulic functions for the final few seconds, potentially giving him more control throughout the controlled ditching.
These are only two of the many such scenarios that are possible. Once all this evidence that proves there was a controlled ditching is widely accepted, interested people will most certainly present other potential scenarios.
The DST group misinterpreted the data from handshake #7. The following describes (in very basic terms) how they concluded MH370 had entered a steep descent.
First, based on data they had for previous handshakes, they projected what the data should look like for handshake #7. They made that handshake #7 projection based on there being no alteration in the airplane’s flight path (the airplane continuing to fly straight and level from handshake #6 to handshake #7).
Then, they compared their projected data to the actual data they got for handshake #7. The data they got for handshake #7 was significantly different from what they had projected, so they knew that after passing handshake #6, there must have been a change in the trajectory of the airplane.
To assess what that change might be, they started with their assumption of an unpiloted airplane. Therefore, they assumed that the airplane did not change direction between handshake #6 and handshake #7. With that assumption, only a steep descent could explain the significant alteration in the data. And of course, that conclusion fit nicely with their assumption that the airplane had run out of fuel. It was totally incorrect, but it all fit nicely together.
There are other potential explanations for the “unexpected” handshake #7 data. Here is one example. At a point somewhere near handshake #6, the pilot entered a deliberate and controlled descending turn. This was part of his positioning for the controlled ditching. A change of direction was not accounted for in the DTS group assumptions. So instead of the data showing an airplane entering into a steep and increasing rate of descent, the data could be reflecting an intentional controlled descent, combined with a change of direction. Again, this is just one potential explanation – there are many others, but we need not speculate further.
I will now refocus back to the issue of whether the official investigation authorities might have unduly influenced the work of the DST group. It is interesting to note that the DST group did not independently develop an end of flight scenario based solely on the satellite data. In fact, it would have been impossible for them to do that. There was simply not enough data available.
It was the ATSB investigators who gave the DST group the end of flight scenario they were to use. In constructing the full end of flight scenario, and in calculating their search zones, the DST group was instructed by the official investigation to assume that MH370 was an unpiloted airplane. They were told that both engines had failed, first one, and then the other.
They were instructed to incorporate the results of the simulator testing into their calculations. The simulator testing informed them that the airplane would react to the asymmetric engine failures by entering an out-of-control steep descent. They were told what rates of descent the airplane would achieve, and how long the airplane would stay in the air, and how far it could travel.
It was with this pre-conditioning and instruction from the official investigation that the DST group interpreted the data at handshake #7. If they looked for other scenarios that would fit with the data, any such alternate scenarios were dismissed. That is unfortunate, but understandable, given that they were instructed to look at only one scenario. It seems that the DST group’s task in assessing the end of the flight was simply to determine how out-of-control the airplane was during its dive into the ocean.
Whether deliberate or not, the interactions between the official investigation and the DST group resembled a mutually beneficial support circle. The DST group was able to prove the robustness of their science when they were able to produce, from unbiased “hard data”, a result that confirmed exactly what the official investigation says actually happened. The official investigation benefited by being proven correct through an independent scientific process that used “hard data”.
The problem, of course, is that all this was based on misinformation. They had used wrong assumptions about an unpiloted airplane, and their deficient wreckage examination gave them an incorrect flap position. Nevertheless, their “findings” provided good cover for the ongoing investigation and search efforts.
Having been in this business for so many years, I find it difficult to believe that some members of the investigation team were not aware of the shortcomings in the evidence being presented by the official investigation.
If some investigators were aware, and if they made their concerns known to the official investigation, their concerns were dismissed or ignored. When you are part of an investigation team, you are expected to express your opinions only within the investigation. You are to argue your positions privately, and never publicly.
As a team member, you agree that only one voice speaks publicly, and that voice belongs to the official investigation. It seems like everyone on the MH370 team has so far lived up to this commitment. Perhaps this book will encourage some members of the team to come forward and reveal what was happening in the background.
In this case, and at this stage where the official investigation has basically run its course, I believe that investigators who were part of the official investigation should abandon the normal protocols for information release. I believe it would be a service to aviation safety if they chose to speak up.
I will say again that I have great respect for the work done by the DST group, and by others who assisted them. Their expertise is obvious, and they did exceptional work to determine a basic track line for MH370. Their dedication is to be commended. That does not take away from their contribution, perhaps through no fault of their own, to the incorrect conclusions reached by the official investigation.
In January 2018, a privately owned company launched a second search for the wreckage of MH370. During that search, the government of Malaysia issued what they called Operational Search Updates. The following is a quote from the background information included with these updates, “On 10 January 2018, the Government of Malaysia entered an Agreement for the search of MH370 with Ocean Infinity Limited”.
The agreement stipulated that Ocean Infinity Limited would undertake a search operation to locate MH370 within a priority search area of 25,000 square kilometers in the southern Indian Ocean. The agreement was based on the principle of “no cure, no fee”, and was to be completed within a timeframe of 90 search days.
The agreement identified three distinct search areas within the 25,000 square kilometer search area, the first being a primary search area of 5,000 square kilometers, the second being 10,000 square kilometers of secondary search area, and the third being another 10,000 square kilometers of tertiary search area.
There was also a fourth search area defined in the agreement, that being a supplementary search area beyond the 25,000 square kilometers. The amount of payment for finding the wreckage, or the flight data recorders, was to increase as follows: 20 million USD if found in the primary search area, 30 million if found in the secondary search area, 50 million if found in the tertiary search area, and 70 million if found outside the 25,000 square kilometer area.
Australia committed its continued support for the new search effort by supplying technical support to the Malaysian Government. Accordingly, Australia supplied all the data related to the original search that had been suspended in January 2017.
Other support supplied by Australia included detailed analysis work on potential wreckage locations completed by the Commonwealth Scientific and Industrial Research Organization (CSIRO). Their work on this project was commissioned by, and funded by, the ATSB. CSIRO’s primary contribution was to help estimate the location of the accident site. They did this by using sophisticated drift modelling, and ocean drift experimentation using replicated wreckage pieces.
It is interesting to note that a major factor influencing the results obtained by CSIRO was that their drift modelling and research started with the assumption of there being a significant amount of floating debris, and a large ocean-bottom debris field, each caused by the breakup of the airplane after a high-speed diving crash. We know that this assumption was not accurate. We know that the airplane remained basically intact, and sank to the bottom largely intact.
The specific 25,000 square kilometer area for the new search was originally identified at a meeting, called First Principles Review, held in Canberra, Australia from 2 to 4 November 2016, less than three months prior to when the official search would be halted. The stated purpose of the meeting was to reassess and validate existing evidence, and to identify any new analysis that may assist in identifying the location of the missing aircraft.
The ATSB’s report on that meeting (dated 20 December 2016) stated, “there were representatives at the meeting from all of the organisations participating in the Search Strategy Working Group including Australia’s Defence Science and Technology Group (DST Group), Boeing, Thales, Inmarsat, the National Transportation Safety Board of the US, the Air Accidents Investigation Branch of the UK and the Department of Civil Aviation, Malaysia. In addition, there were representatives from the Commonwealth Scientific and Industrial Research Organisation (CSIRO), Geoscience Australia, Curtin University, Malaysia Airlines and the People’s Republic of China”.
The meeting was held with the knowledge that their ongoing search for the wreckage in the southern Indian Ocean was winding down. It was increasingly apparent that the wreckage would not be found. They were confident in their search methods, so they were able to conclude that the wreckage was not within their identified 120,000 square kilometer search zone. They used that knowledge, along with updated information from other sources, to define the most probable unsearched area where the wreckage might be located, which was the 25,000 square kilometer area used for the Malaysian Government’s agreement with Ocean Infinity Limited.
In their deliberations at this First Principles Review, the participants continued to rely on the inaccurate conclusions about how MH370 flew to the southern Indian Ocean. Their deliberations and calculations continued to be tainted by the misguided assumptions that the airplane was unpiloted, and that the engines failed, and that MH370 had experienced a high-speed diving crash.
The ATSB’s report on the First Principles Review confirms that their rationale for choosing the new 25,000 square kilometer search area relied on the following: 1) the assumption of an unpiloted airplane – that assumption was built into the study of the satellite communications metadata that was used to determine the potential flight path; 2) the results of the simulator work, where they assessed the parameters of an impact zone based on how far the airplane could travel in an out-of-control high-speed dive, following a double engine failure; and 3) their (inaccurate) analysis of the recovered wreckage, where they found that the flaps were retracted, and that the airplane had crashed in a high-speed dive.
By starting with these basic inaccurate assumptions, it is not surprising that the chosen 25,000 square kilometer zone of ocean bottom did not contain the wreckage of MH370. This new search had no greater chance of success than did the original search.
Similar to what had happened during the original search, some of the contributors who had an influence on choosing the new search location created very high expectations for success. In their public statements, they indicated they were able to determine the location of the wreckage with high precision.
As discussed earlier, it is never appropriate for people connected to this type of activity to create such high expectations by predicting a positive and speedy outcome. There should never be a promise of results, only a promise of effort.