More data have been generated, collected and processed by the human race in the last 36 months than during the previous 3,000 years. Further, every day of every month the worldwide flood of information on every conceivable subject increases exponentially.

We are living in the era of “Big Data,” and every discipline, including aviation, is involved. In the digital age, enabled by cyber technology and the Internet, ones and zeros dominate.

Every activity in which we engage, every purchase on the Internet, every phone call, text message or email we make or send, leaves digital data trails. Tracking these and processing the information they contain forms databases held in high value, especially by the business community, and thus increasingly, Big Data is evolving as the currency of the digital age.

Aviation has always been at the forefront of data collection, especially during the civil component’s formative years when it was obvious that if the nascent airline industry were to grow and prosper, it had to make operational safety its highest priority. And assessing safety — to have a gauge for measuring performance and identifying potentially dangerous trends — requires data.

Early on, manually gathering, formatting and reporting operational and maintenance data was a laborious and time-consuming process, but cybernetics, automation and the digital revolution beginning in the 1960s not only simplified the job but also enabled the reaping of ever larger amounts of information.

By the early 1980s, even the aircraft themselves were recording performance, maintenance and trending data for later downloading into storage devices or laptop computers. Less than a decade later, many were automatically transmitting hundreds of parameters of near-real-time performance data to maintenance departments or trend-monitoring companies while in flight.

(I remember watching at a monitor at United Airlines’ San Francisco maintenance base as engine performance numbers streamed in from a Boeing 747-400 cruising over the Western Pacific at FL 410 on a revenue flight between Japan’s Narita International Airport (RJAA) and SFO. The data, showing everything that was happening within the four powerplants, were collected electronically, uplinked to a comm satellite in geostationary orbit, then beamed down to the SFO facility. That was 20 years ago.)

At the same time, air navigation service providers (ANSPs) were implementing computer software that automatically recorded gross navigation errors (GNEs) and altitude deviations — often to the chagrin of offending flight crews but nevertheless to the ultimate benefit of the overall safety system. As the technology matured, the data wave turned into a tsunami.

What Does It All Mean?

So, as Russell Lawton, safety director at the Air Charter Safety Foundation (ACSF), quipped in an email exchange with BCA, “What does it all mean?” Indeed, to be useful, data must be organized and analyzed.

All aviation risk-management programs and safety management systems (SMSs), including the International Business Aviation Council’s (IBAC) International Standard for Business Aircraft Operations (IS-BAO), depend on accurate flows of data in order to reveal patterns that could presage trouble. It is submitted and collected in two forms: digital and textual. An example of the former would be data downloaded either manually or wirelessly from aircraft, while an example of the latter would be incidents or notices of potential problem areas submitted to the safety-reporting programs.

“What we are doing . . . is looking at the data to verify how we’re flying the aircraft,” explained Frank Raymond, an aviation safety officer for a major western U.S. business flight department that has participated for five years in an FAA-approved Flight Operational Quality Assurance (FOQA) program which he manages.

In a prior position, Raymond honed his risk-aversion skills at an airline where he oversaw an Aviation Safety Action Program (ASAP), another risk-management scheme detailed further on. Until recently, he also served as chairman of the steering committee for the C-FOQA Centerline user group of 70+ business aviation operators that contract with General Electric Flight Efficiency Services to analyze their flight data. The ensemble collectively operates a total of 140 business aircraft.

C-FOQA Centerline is “a conduit to GE in that regard,” Raymond said, “collaborating in terms of projects that would be relevant to the overall group.” And the relationship yields one other benefit: “Because we are a member of the C-FOQA group, we are benchmarked against the other operator members. With benchmarking, we can then compare our performance with more than 70 other operators.”

While IS-BAO addresses quality and safety management of the entire department, FOQA focuses solely on flight operations, harvesting digital data from aircraft FDRs and flight computers. “We’ve used IS-BAO to describe how we do business in all parts of our department but rely on FOQA to monitor the flight operations component,” he said. “It feeds into our safety program, which, in turn, is part of the IS-BAO program.”

Another of Raymond’s responsibilities is “gatekeeper” for his company’s FOQA program, protecting the security of data and its connection to flight crewmembers and maintenance staff. This is significant, since the effectiveness and integrity of FOQA and similar safety-reporting programs are based on immunity from FAA action as long as crews and techs are not engaged in patently illegal activity. In order for participants to be forthcoming, a non-punitive atmosphere must prevail.

FOQA was conceived by British Airways as a tool to study stabilized approaches and was subsequently adopted by other airlines and, eventually, the FAA as a completely voluntary data-collection program. As it was non-punitive, it quickly gained in popularity. While generally oriented toward airlines, FOQA can also serve business and general aviation operations, e.g., Raymond’s flight department and its counterparts in the GE Flight Efficiency Services fold. To assist U.S.-registered operators in establishing a FOQA program, the FAA issued Advisory Circular 120-82, dated April 12, 2004, for “guidance on one means, but not necessarily the only means,” of developing, implementing and operating a program acceptable to the agency.

AC 120-82 states that the intention of FOQA is to increase operational safety by allowing operators and pilots “to share de-identified aggregate information with the FAA” so the aviation authority can monitor operational trends and address “risk issues.” The FAA views the program as a “partnership” between employers (the airlines or operators), flight crews and itself “to identify and reduce or eliminate safety risks, as well as minimize deviations from the regulations.” Further, the “cornerstone” of the program “is the understanding that aggregate data that is provided to the FAA will be kept confidential” and the identity of reporting pilots or operators will be anonymous “as allowed by law.” Thus, under 14 CFR Part 193, submissions to the program will be protected as “voluntarily submitted safety-related data.” However, to obtain assurance of protection “from the use by the FAA of FOQA information for enforcement purposes,” operators must obtain FAA approval of their programs, hence the 78 pages of guidelines supplied in AC 120-82.

So essentially, FOQA program approval cements an understanding between the operator and the feds that pilots can submit safety concerns to the program without fear of reprisal or punishment. Operators and employees can then be confident that, in this sense at least, their data is “secure.” The premise for this protection is that the greater good of being able to identify unsafe trends from the data submitted outweighs punishment, or as Raymond put it, “We are looking for trends in how we fly to ensure that our employees are operating the aircraft within set standards and per their training.”

Data used in FOQA programs are the same types stored by digital flight data recorders (FDRs) for accident-investigation purposes but can be downloaded frequently from quick-access recorders. “The FDRs can record thousands of parameters of data,” Raymond pointed out. “The pieces we’re retrieving aren’t that extensive in detail but sufficient to understand what the aircraft were doing, how high and fast they were flying, and so forth.

“Data from our aircraft are downloaded once a month and forwarded to GE for analysis and trending,” he continued. “As part of GE’s analysis, we get quarterly and annual reports that show what we’re doing on a trending basis, and these tell whether we are operating in compliance with our department SOPs, with the manufacturers, and in accordance with how we’ve been trained. In our operation, we have adopted the Flight Safety Foundation’s standard for a stabilized approach and developed our program around the following parameters: correct flight path, aircraft configured for landing, speed within range, appropriate power setting to maintain correct descent rate, and establish an approach ‘gate’ 1,000 ft. before touchdown.”

Tools for Building a Safety Culture

ASAP, mentioned earlier, was launched by the FAA in 1997 in conjunction with participating airlines. Like FOQA, it is a confidential, voluntary reporting program designed so that industry participants are assured security in reporting incidents; i.e., it is non-punitive. A separate program, the Air Traffic Safety Action Program (ATSAP), has been tailored for controllers with the same objectives and protections.

The ACSF’s Lawton described ASAP “as a hands-on safety process whereby people come forward and report events they have experienced in exchange for determining root causes and developing corrective action to mitigate hazards, a more formal way of looking at safety because the whole idea is that it is intended to get to the cause and not assign blame. When an issue is identified, we examine how we are currently controlling the risk or whether procedures and training should be revised.”

ASAP “owes its origins to the Part 121 world,” Lawton said, “where it has subsequently established a proven track record in safety enhancement.” However, since its founding in 2007, the ACSF has been approved as a neutral third-party ASAP program manager for Part 135 and 91K operators and now Part 91 flight departments, as well. “Whatever works most effectively getting to the root of the event and is going to make the [reporter] forthcoming is the program’s objective. The program can identify trends and patterns, symptoms of larger problems that need to be addressed like, for example, noting that crews are having problems with automation. And it can even report problems back to the OEMs.”

Unlike FOQA, which focuses on stabilized approach parameters, ASAP addresses a broader purview of operations, allowing anyone involved in an operation to report incidents or anything that could indicate a potential safety problem. “We now have the capability to do something with this information,” Lawton said. “FOQA and ASAP are complementary tools; the former will tell you what happened but not why — you need ASAP for that.”

When an ASAP report is received, it is screened against what the FAA terms the “big five” caveats that could disqualify it: The incident occurred due to criminal activity, substance abuse, use of a controlled substance, alcohol abuse or intentional falsification.

If an Event Review Committee (ERC) determines that the report qualifies, it is investigated through a formal process. There are two categories of reports: sole source and non-sole source.

“A sole-source report is one where no one would know what happened if the reporter had not come forward,” Lawton said. A non-sole source report is one in which the FAA has knowledge of the incident, e.g., altitude busts, speed deviations, nav errors or things controllers would see and the ATC system would record. (A software tool called Air Traffic Quality Assurance [ATQA] records ATC incidents and identifies pilot-caused events, thus defining a non-sole source report, as ATC already would have a record of the incident.)

An ERC would include representatives of the FAA, the airline or company employing the reporter, and an employee of the operator (not the reporter). After interviews with the reporter (or flight crew), it then develops corrective action for either the individual or the company involved. “All decisions are reached by consensus,” Lawton observed, “a very important point. Everyone has to come to agreement on the corrective action, which is reviewed periodically to ensure implementation and effectiveness. The objective is to get the information needed to identify a problem and develop corrective action to avoid or solve it.”

ASAP is an “excellent tool for building a safety culture,” Lawton claimed, “because it puts in black and white that you get protection and you are helping to advance safety. Thus people will get more comfortable in sharing what is happening to them. It encourages a dialogue that otherwise might not have occurred. The idea is that this information is used for constructive purposes and not to lay blame.”

There are two ways to destroy a safety reporting program, Lawton maintained. “One is to ‘burn’ the reporter and the other is to do nothing with the information. The way ASAP is set up is that you get a secure database to host the [reported] incidents that allows reviewers to access specific airports or regions to review them. At the ACSF, once a report is investigated and de-identified, it goes into a dedicated server so operators can access the information for learning purposes. Your company always owns its own data, not the FAA, and that’s why you get your own secure server.”

The majority of Part 121 operators in the U.S. are participating in ASAPs, with more than 90,000 reports submitted last year, of which 90% were sole source. The number one event reported across ASAP is altitude deviations.

It’s Not Only Nice to Share, It Also Can Save Lives

As the data roll in and these individual risk-management databases grow, their potential usefulness cannot be achieved unless the information they contain is shared. Enter the FAA’s Aviation Safety Information Analysis and Sharing (ASIAS) program, launched in 2007 to aggregate and disseminate data from FOQA, ASAP, ATSAP programs and other sources. Operated under contract to the FAA by the MITRE Corp., ASAIS produces reports and issues alerts on trouble spots and negative trends revealed by the contributing programs.

“ASIAS is a collaboration between the FAA and the aviation industry,” said Randy McGuire, department head, aviation safety analysis, at MITRE’s McLean, Virginia, laboratory. The lab is one of seven in the U.S. performing research for the federal government in multiple fields including aviation.

“It was started with the intent of collecting and protecting safety data that U.S.-based airlines shared with MITRE as an objective third party,” McGuire explained. “A few years ago we began sharing with general aviation operators.” Currently, there are 45 airlines and 20 FAR Parts 91, 91K and 135 business aviation operators in the program. ASIAS has also been extended to Part 145 repair stations and Part 141 flight training schools.

Operators in the program represent one aircraft each to entire fleets numbering in the hundreds of aircraft, with about 850 aircraft represented for general aviation. Additionally, there are training fleets for flight schools and universities (a growing area), and piston operators using their aircraft for business travel and for which they’re developing metrics as well.

ASIAS is not a data source, McGuire explained. The program is an aggregation of multiple data sources, including the Aviation Safety Reporting System (ASRS), FOQA, ASAP, ATSAP, and recorded radar and weather data. The data processed are both text-based (the safety reporting programs) and digitally based (like FOQA harvests from digital aircraft reporting systems and FDRs). Text-based data are analyzed with text-mining software that “reads,” “translates” and categorizes an event as one of several that safety reporting programs and ASIAS track. “All the data are automatically processed, and we receive them in electronic formats,” McGuire said. “So every day the data flow in here. . . .”

As an example of how the aggregation process works, digital data from an aircraft are collected by an operator under FOQA and sent on to MITRE to be added — i.e., aggregated — with data from other operators, de-identified and kept in a secure location. Every two months the archive is pulled and computer metrics based on the “snapshot” are created or updated and the archive is refreshed with new information. Finally, “dashboards,” or graphical representations of the pulled data described by month, are created and uploaded to a portal, a secure website, to which each operator in the program has access.

“They can log in and view the risks we have identified every two months and actually see the results,” McGuire said. “The airline operators will see benchmarks of their safety rate against the larger group, but at the present time, only them. GA operators see the aggregate of everybody else and can compare that to their own statistics they keep internally. These sites are secure and represent an incentive for operators to participate.”

Examining the aggregated data, analysts can decipher trend lines. “So if something is deteriorating, we will see the trend line increase,” McGuire said. “One way of making the trend lines decrease is when the [ASAP] General Aviation Joint Steering Committees or the Commercial Aviation Safety Teams [CAST] at the airlines put forward risk mitigations, or suggestions to the operators on how to reduce the risks they’re seeing in the data. Once these are implemented, ASIAS then determines whether they are successful using the same process employed to identify the risk in the first place: studying the data, a process that can take several years.”

The value of this system, McGuire pointed out, “is when we see something we have not seen before that becomes a new risk to address before it becomes an accident. It enables us to be proactive, to see things that could identify dangerous trends. This might be something we haven’t seen before or something we’ve been tracking that has changed or shown up in new places.”

Since ASIAS has opened the program to general aviation input, Jeff Middelman, the program’s general aviation lead, said, “we have agreements with operators that do the same type of analysis from a uniquely general aviation or business aviation perspective.”

Working with commercial or general aviation operators, “We come up with topics they are interested in tracking,” he explained. “To that end, we look for different metrics like FOQA that we compute and update on a regular basis to our stakeholders, and here we can see the trends over time at different airports and for different aircraft.”

This is the aggregated form of all operators’ data put together that can reveal developing trends. “And of course, all the data are de-identified,” Middelman said. “There could be tens of thousands of text-based reports plus the digital data from programs like FOQA. Textual reports are used only when a crewmember reports voluntarily.”

The digital data by now represent millions of flights, e.g., 15 million for commercial operations alone covering a three-year period, as the data are stored for 36 months, then deleted. “So you wind up with a ‘rolling total,’” Middelman said. And in terms of the textual reports, “Just for the airline component, we are collecting and archiving 150,000 text reports every three years. On the general aviation side, with only 20 operators in the program, the total is smaller: Currently, we have 20,000 FOQA and 13,000 ASAP reports.”

How It’s Paying Off

So, have trends revealed in the data led to interventions that have made operations in the nation’s airspace safer? According to McGuire, based on data from airline operators, there have been 19 safety enhancements implemented by the CAST in the past eight years. “On the general aviation side,” he said, “we have not collected enough data to ascertain trends. We are working on that. We need more general aviation operators to join the program to increase the size of the database. We need a lot of operators to get a lot of data over time.”

In order to join ASIAS, an operator must be either a FOQA or ASAP data contributor and have a safety program in place. As a general aviation operator, McGuire said, “You can use an SMS provider, registration in IS-BAO or internal programs that do not have be FAA approved. On the other hand, airline programs do have to have an FAA approval. There is no cost to join. You receive access to the portal, study results done previously and invitations to the semiannual FAA InfoShare conferences held around the country at different locations. You can participate in special studies of the trends, where they are occurring. You can request a study. You can help in developing a safety enhancement or mitigation.” To get the process started, visit http://www.asias.aero.

As an example of how an operator in the program can plumb ASIAS safety data, McGuire cited unstable approaches as a topic. “We can look at unstable approaches at any airport for which we have data. General aviation flies to a lot small airports. Pick one, and the user can log in and go to that airport and determine what the unstable approach rate is. It’s possible to see it for each runway, too, so the rate of unstable approaches at the field may be driven by one particular runway.” At least two operators participating in ASIAS have to be using an airport for the program to have data on it.

Once in the program, any use of the data has to be approved by the ASIAS executive board to ensure protection. “The program has strict governance in place to protect the data, who can see it, how it is used, to ensure operators are comfortable in providing the data,” McGuire said. “The steering committees help oversee how the data is used and how it might be initiated. Industry is always part of that — operators, labor unions, industry groups like the NBAA, and the OEMs, [the last which] often assist with data. It is a totally collaborative process with full transparency. It’s all about building trust, which has taken a lot of time. On the general aviation side, we need a lot more operators to get a lot more data. It took years to get the airlines into it.”

Observed Frank Raymond, “The airlines are feeding data from thousands of flights into these programs daily. Meanwhile, business aviation may only be submitting a small amount of data, relatively speaking, but every bit helps to raise the bar and give us a bigger data set to look at. So everybody benefits in the end.”