A version of this article appears in the March 17 edition of Aviation Week & Space Technology.

The story of Malaysia Airlines Flight MH370 reads like bad fiction—a perfectly safe aircraft disappears from radar on a scheduled flight, and the extensive international search-and-rescue effort that ensues finds few clues. It is not fiction, but it should be in light of the technology to monitor aircraft and the massive amounts of operational data available.

On March 8, the aircraft, a Boeing 777-200ER registered 9M-MRO, operated the daily MH370 service from Kuala Lumpur to Beijing. Onboard were 227 passengers and 12 crewmembers. About 40 min. into the flight—around 1:30 a.m. local time—radar and radio contact was lost over the Gulf of Thailand, 90 nm off the Malaysian coast, near Kota Bharu. The aircraft is believed to have been on airway R208 heading for the IGARI waypoint, about to be handed over to Vietnamese air traffic control; it never checked in.

While the initial search focused on the expected flightpath of the aircraft between Malaysia and Vietnam heading principally north, the effort has been broadened. Ships, aircraft and helicopters were dispatched west and east of Malaysia.

According to a Wall Street Journal report, the aircraft kept sending signals without content through the Aircraft Communications Addressing and Reporting System (Acars) for 4 hr. after the last radar and radio contact, which would indicate that the 777’s engines continued to run throughout that period. This led to speculation about a possible hijacking. Boeing and engine manufacturer Rolls-Royce had no comment.

Malaysia Airlines CEO Ahmad Jauhari Yahya says the last Acars message was received at 1:07 a.m., well before radar contact was lost.

The airline took delivery of 9M-MRO in 2002; the aircraft accumulated 53,465 hr. and 7,525 cycles since then. It underwent scheduled maintenance on Feb. 23, 12 days before the March 8 flight.

Secondary surveillance radar (which interrogates the aircraft transponder for identity and position) only works within range of the radar. Automatic dependent surveillance-broadcast (ADS-B) also relies on aircraft transponders, which can be turned off manually. But had this been the case, primary radar still should have detected the aircraft. ADS-B continuously sends key locator information and works beyond radar coverage, although for now, only within range of a ground station. However, data can be received by other aircraft in the vicinity that are equipped to do so.

Surveillance over the South China Sea is improving, thanks to ADS-B systems that have been installed by several countries in the region. Singapore, Vietnam and Indonesia have such coverage and have agreed to share data. As of December 2013, aircraft are required to be equipped for ADS-B above Flight Level 290 on certain trunk routes that overfly the South China Sea—although those in question are farther to the east than MH370’s scheduled route.

According to International Civil Aviation Organization (ICAO) documents, Malaysia is also implementing ADS-B surveillance to improve coverage of certain air routes that do not have complete radar coverage. This entails installing two ADS-B ground stations at radar sites in Pulau Langkawi and Genting Highland. Installation and testing are slated for 2014-16. Data-sharing with other countries and mandatory aircraft equipage on some routes are planned, ICAO states.

Eventually all commercial aircraft will stream live-status data directly to operation control centers, offering real-time situational awareness that is all but nonexistent today. The key question is whether operators will opt for their datasets and upgrades based on return on investment (ROI)—as they do now—or in response to a mandate issued to prevent another case like MH370.

Malaysia, like most other airlines, has not invested in a system that supplements standard aircraft-status-reporting channels such as air traffic control or Acars—a data link using VHF, HF or satcom over oceanic airspace whose messages are routed via ground stations to the end users.

Research led by the French accident investigation bureau (BEA), following the 2009 crash of Air France Flight 447 (AF447), underscored that feasible supplemental systems exist and can perform safety-related tasks such as tracking aircraft and automatically offloading flight-data recorder (FDR) parameters that go beyond the inflight engine-monitoring function used by many airlines. But the few operators that own them are primarily motivated by day-to-day operational benefits. “Nobody is going to purchase those services without a business case and a return on investment,” says Matt Bradley, president of Flyht, which provides flight-data streaming on about 300 aircraft flown by 45 operators.

Today, that ROI comes from monitoring variables such as engine-surge data to head off unplanned repairs, or tracking maintenance faults so ground stations have what they need to keep inbound aircraft on schedule. It is often ancillary benefits from these systems, rather than data derived from safety-specific ones, that provide accident investigators with key insight.

The first clues about AF447’s fate came via 24 automatic Acars messages relayed through the A330’s Air Traffic Information Management System between the time of its last position report and its impact with the Atlantic Ocean.

Investigators probing the September 2010 crash of UPS Flight 6 near Dubai relied on a combination of Acars messages and data from Boeing’s Airplane Health Management (AHM) real-time monitoring service to determine where the fatal onboard fire started, and not from the charred wreckage of the 747-400F.

UPS’s AHM strategy underscores the reason that, even when carriers eye the bottom line, such systems are not more widely adopted. The carrier uses AHM for its 747-400s and MD-11s. Their long stage lengths give the carrier a better chance to act on fault messages and to have parts ready at a ground station, cutting aircraft downtime. Reliability is another factor. UPS’s workhorse Boeing 757s and 767s do not produce enough schedule-crippling faults to justify the cost of AHM.

Acars helped shed light on some aspects of the AF447 crash, but not on locating the wreckage. BEA in 2010 created a working group to examine how best to pinpoint wreckage location and to alert others of a problem before an aircraft disappears from radar screens. The group ran simulations using 689 sets of actual flight data, including 44 accidents and 28 incidents. It established parameters, such as unusual attitude, speed anomalies and warnings from onboard systems like TCAS (traffic-alert and collision-avoidance system) that would trigger an automatic response. Tested under binary (true/false) logic rules, the parameters detected 67 of the 68 accidents, and registered two normal flights as possible accidents. Most important, warning times were sufficient to send out relevant FDR data. Applying fuzzy set theory to the parameters, which gives them weight, detected all of the accidents with no nuisance alerts.

BEA then collaborated with Inmarsat to simulate each accident in 597 locations worldwide. Analysis showed that pre-impact transmission would have been possible in 85% of the combinations. In 82%, the corresponding wreckage search zone would be within a 4-nm radius. Trials with the globe-covering Iridium satellite network were similarly encouraging. “This study proves that developing reliable emergency detection criteria is achievable,” the working group report stated.

BEA’s research, which led it to recommend additional study into live data-streaming, included a demonstration of Flyht’s emergency mode triggered by either manual input from the air or the ground, or predefined parameters. The monthly cost is approximately $1,500 per aircraft, depending on service levels. Coverage is no issue.

However, data-streaming has some drawbacks. Bandwidth costs tip the scales toward event-triggered systems, but these rely on a precursor, such as an aircraft upset, which makes them useless in events such as rapid decompression or inflight explosions.

Ka-band, which powers onboard Wi-Fi on JetBlue Airways and United Airlines aircraft, offers promise. It costs at least 70% less than the more established Ku-band and offers much more bandwidth. While ancillary revenue from inflight entertainment is driving today’s equipage decisions, Ka-band service provider LiveTV is convinced airlines will eventually equip for operational benefits; in-cabin revenue will be added value.

Any service based on inflight Internet has a coverage challenge, and Ka-band, being the newest, has the biggest. While service providers expect coverage gaps to be filled and roaming agreements to be struck, they acknowledge that a few satellite launches and business agreements are needed to bridge that vision and today’s reality.

Tests to prove the viability of upgrading a standard-equipped flight deck for satellite-based connectivity were conducted onboard Boeing’s 737-800 EcoDemonstrator in 2012. A version with higher bandwidth capability is due to be tested this year on a 787 development aircraft, which will be used as the company’s second EcoDemonstrator (see page 46). Initial tests in 2012 evaluated an iPad-hosted flight trajectory optimization app which funneled real-time weather data to flight-deck crews. The data were the same as those normally transmitted to flight crews over the continental U.S. But in the case of the EcoDemonstrator, data were sent via satellite to simulate connectivity over oceanic airspace, where live-weather data is currently not available.

Boeing is studying the possible evaluation of a Ku-band satellite system on the 787 EcoDemonstrator.

- Sean Broderick in Washington, Jens Flottau in Frankfurt, Guy Norris in Los Angeles, Adrian Schofield in Auckland