Controlled evidence-assurance research summary

Malaysia Airlines Flight MH370

Aperture evidence assurance applied to a global, high-uncertainty public-source investigation.

MH370 is used here as a practical demonstration of what Aperture does when a decision record contains official findings, technical models, physical evidence, incomplete primary records, competing explanations, public speculation and material unanswered questions.

Research state: Version 0.6 research closure. Evidence cut-off: 7 August 2026. Independent specialist review: pending.

International access: concise reader guides are available in English, Bahasa Melayu, Simplified Chinese, French, Spanish, German, Arabic and Japanese.

Case type: Independent public-source demonstrationClient status: Not a client engagementReview status: Controlled research summary / public PDF pendingSite review: 16 August 2026
Share this case

Public PDF status

Full public PDF not yet released.

The underlying MH370 research package remains on publication hold pending independent specialist review, institutional right of reply and final release assurance. This page is a public capability summary only. No Aperture PDF is released from this page until those gates are completed.

Publication control: a preview or controlled research file must not be presented as a final public PDF.

Decision structure

What this case was really testing

The claim

Competing explanations and public claims about the disappearance needed disciplined separation.

The uncertainty

A large body of reporting did not eliminate major evidentiary gaps.

What had to be tested

Source origin, chronology, official material, alternative explanations and unresolved evidence gaps.

Decision relevance

Which propositions could responsibly be treated as supported and which remained unresolved.

Document release control

The research record exists, but the public document set is intentionally withheld.

The underlying MH370 research package remains on publication hold pending independent specialist review, institutional right of reply and final release assurance. This page is a public capability summary only. No Aperture PDF is released from this page until those gates are completed.

When the release gates are complete, this page can expose the approved full PDF directly. Until then, the absence of a download is deliberate.

Why this example matters

This is what evidence assurance looks like when the information is abundant but certainty is not.

A client does not need thousands of pages of material simply repeated back to them. The problem is to identify which propositions are established, which depend on models or assumptions, which claims recycle the same source, which contradictions survive testing, which evidence remains unavailable and what level of reliance is responsible.

The MH370 record makes those controls visible. It shows how Aperture separates the broad reconstruction from unresolved questions about initiating cause, controller identity, intent, motive and exact wreckage location.

What a client can inspect here

  • the controlling investigative narrative
  • claims-and-evidence matrices
  • a master chronology
  • source and evidence registers
  • technical annexes and adversarial analysis
  • evidence-gap and specialist-review controls
  • decision and reliance records
  • machine-readable registers
  • international reader guides in eight languages

The decision question

What does the accessible evidence establish, and where must reliance stop?

Establish the broad record

Reconstruct the flight, surveillance loss, radar phase, satellite signalling, debris recovery and search history from traceable source families.

Challenge the narrative

Test technical interpretation, independent corroboration, public claim mutation and competing causal explanations rather than rewarding repetition or confidence.

Define the reliance boundary

Separate high-confidence broad findings from model-dependent propositions, unresolved causal questions and evidence that only future recovery or specialist work can resolve.

Decision-grade result in the supplied record

What the strongest evidence supports

01

Broad flight reconstruction

The supplied dossier supports normal initial departure, later loss of normal secondary-surveillance identification, westward diversion and continued multi-hour flight.

02

Southern route and loss region

The supplied research classifies the southern route and ultimate loss in the southern Indian Ocean as very strongly supported, reinforced by confirmed physical debris.

03

18:25 SDU event

Restoration of SDU power after a material interruption is strongly supported. The historical cause of the interruption remains indeterminate.

04

00:19 sequence

Fuel exhaustion followed by APU-supplied electrical restoration is the best-supported public explanation in the supplied record, subject to Boeing 777 systems specialist review.

05

Final BFO evidence

The final measurements are treated as strongly consistent with a substantial and rapidly changing descent component, but not as direct proof of one exact descent rate, attitude or control state.

06

Cause and actor boundary

The supplied record does not establish initiating cause, controller identity, intent, motive or exact wreckage coordinate. Competing broad causal families remain unresolved.

These statements summarise the supplied v0.6 research-closure record. They are not presented as an official accident-investigation finding or as independently validated specialist conclusions.

Contradiction testing

The case shows why repeated claims still have to be tested.

Exact final descent rate

The record rejects treating final BFO values as a direct recorder of one exact vertical descent rate.

18:25 proves deliberate action

The technical power interruption/restoration proposition is separated from the unresolved historical cause.

“2.4 tonnes of lithium batteries”

The v0.6 addendum records 221 kg of lithium-ion batteries within a broader 2,453 kg Motorola consignment and treats the common formulation as materially misleading.

Simulator proves rehearsal

The official forensic material is treated as relevant circumstantial evidence, not proof of one exact rehearsal of the later route.

One searched area equals total exclusion

The dossier distinguishes higher-confidence sonar coverage from lower-confidence and data-gap areas rather than treating all searched seabed as uniform 100% exclusion.

One adversarial paper settles the issue

The alternative controlled-descent model is retained as a serious challenge without being allowed to displace the wider record merely because it is novel or confident.

Research readiness

Research can be mature without pretending that the case is solved.

The supplied v0.6 readiness report records a score of 89 / 100 and classifies the matter as Ready Subject to Independent Specialist Review. General desk research is closed under a diminishing-value test for the evidence cut-off.

The score does not override material deficiencies. The record still identifies missing primary evidence, specialist reproducibility work, current-search quality-control data, right of reply and external review as limitations on external reliance.

Examples of material gaps

  • main wreckage and exact coordinate
  • FDR and CVR
  • full raw military-radar dataset
  • full original simulator forensic corpus
  • independent SATCOM numerical reproduction
  • Boeing 777 SDU/electrical specialist validation
  • current Ocean Infinity AUV/QC data
  • direct actor and intent evidence

What this solves for a client

The subject can change. The decision problem is familiar.

You have too much information

Aperture converts documents, claims, data and commentary into a controlled evidence architecture rather than a larger pile of reading.

Experts or sources disagree

Competing explanations are kept visible, independent corroboration is separated from genuine corroboration and technical questions are escalated instead of guessed.

You need to know what is safe to rely on

The output distinguishes established facts, supported inferences, model-dependent findings, unresolved questions and the evidence that could change the answer.

Your matter does not need to be as large as MH370. The same controls can be scaled to a counterparty, transaction, disputed claim, chronology, publication, technical question or strategic decision.

Professional boundary

This is an independent public-source research case, not an official accident investigation.

The supplied dossier expressly states that it is not an official accident investigation, legal opinion, engineering certification, accredited forensic examination or substitute for the Malaysian Annex 13 investigation. It does not claim access to non-public evidence.

The source PDFs are published here unchanged. Their original control language remains visible, including the fact that independent specialist validation, institutional right of reply and other release-assurance gates were incomplete when the research package was closed.

Website publication does not convert those unresolved gates into completed specialist validation.

How to use the example

Do not ask whether your matter looks like MH370. Ask whether your decision has the same evidence problem.

If your decision depends on incomplete records, conflicting accounts, technical interpretation, repeated claims, uncertain provenance, source recycling or important unknowns, the MH370 package shows the kind of control Aperture can apply.

INDEPENDENT PUBLIC-SOURCE CASE STUDY - NOT A CLIENT ENGAGEMENT

Aperture Research Works did not represent any party and did not have access to confidential evidence unless expressly stated. This assessment is limited to lawfully available sources reviewed up to the stated evidence cut-off.

Discuss a Matter
V168