Resistance to requiring safety management systems (SMS) and analyzing data from routine flight operations was obvious in comments from some business aircraft operators during April conferences in Canada and the United States. Yet many attendees were intrigued by the prospect of voluntarily resolving familiar but intractable safety issues through these methods — even one concept that also would integrate flight simulator data into SMS.
“Forty years ago … accident prevention was based largely upon having investigators … running out to a debris field, sifting through the wreckage and collecting all the broken parts and pieces [to] literally try to link together … the chain of events, what caused the accident,” said Steve Charbonneau, senior manager of training and standards, Altria, and chairman of the steering committee for C-FOQA Centerline, an industry initiative that collects and analyzes de-identified corporate flight operational quality assurance (C-FOQA) data. “Today, the data that we collect on our airplanes represents that debris field, and we are privileged and fortunate enough through technology to sift through that data to collect the links that identify the risks before accidents happen.”
Charbonneau, also vice chairman of the National Business Aviation Association (NBAA) Safety Committee, was a presenter during Flight Safety Foundation’s Business Aviation Safety Seminar (BASS) in Montreal. “Flight data monitoring has been stigmatized over the years as a tool to single out individual pilots [who] perhaps are not compliant or not performing,” he said. “I want to be very clear that [C-FOQA Centerline] does not promote or support any kind of data gathering or monitoring system in this regard. … In fact, [prohibiting such a use] is written into the language of the [voluntary participants’] service agreement to reinforce that point. Flight data monitoring programs work hand in hand with your SMS as a safety-assurance tool to ensure that your training, standards and risk communication tools are working effectively.”
The program, initially co-sponsored in 2003 by the Foundation and the NBAA Safety Committee, is an industry-led venture for business aviation in North America. “In 2010 and 2011, the program really began to accelerate,” he said. “We had a growing community of [SMS]–compliant operators and IS-BAO-certified operators1 out there that were willing to take a leap, and take a chance with the program. During that time, the program nearly doubled in size with key operators coming on board. … The steering committee … management and leadership of the program is from the user’s group, so this is a grassroots organic program.”
Latest Findings
The program in April published its 2012 aggregate-data report and analysis, updating previous findings published by AeroSafety World (2/13). “We added 17 aircraft in 2012, and we recorded more than 13,000 flights,” Charbonneau said. “The fleet represents a good cross section of airframe types from wide-body Boeings to speedy little [Cessna] Citations. … This data set really is only a small slice of all the operations that are going on out there, and we are actively pursuing strategies to recruit other [FOQA] programs into our aggregate data set.”
His presentation focused mainly on trends in unstable approaches, including insights into this industry segment’s closely related concerns over runway excursions. “When we look at the annual unstable approach rates year-over-year since the launch of the program, we see a shallow decline in rates,” Charbonneau said. “However, when we consider the years of enrollment and the operational experience of operators, we see an obvious reduction in unstable approach–event rates. We can deduce from this … that new operators coming into the program will benefit from robust mentoring, and this is one of our key roles as a steering committee.”
The latest data analysis found that a “disconnected” flow of safety communication to flight crews in summer months seems to explain higher unstable approach rates, and the weather in North America seems to contribute. “It’s clear that visual approaches remain … a challenge for pilots,” Charbonneau said. “Unstable approaches in visual conditions are 20 times more likely to happen than [unstable approaches] in instrument conditions. This is [the case] even when we reduce the tolerance, the gateway to achieve approach stability, from 1,000 ft to 500 ft.”
C-FOQA Centerline recently collaborated with the Foundation in developing safe landing guidelines for participants, then monitoring adherence to the guidelines. “[The latter part of] 2012 was the first year that [C-FOQA Centerline] decided to take a deeper look at landing stability and understand the effects of landing performance with regard to safety,” he said. “Performance is monitored in three critical phases: threshold crossing, touchdown and rollout. Any single event will trigger an unstabilized-landing caution.” No significant changes have been noted.
Nevertheless, data from 2012 piqued the steering committee’s curiosity about practical implications. “Firstly, four of the top five [causal] factors of unstable approach events are related to high energy states,” Charbonneau said. “[Some crews were] either high or fast on approach, [a state] developing into late gear extensions and high rates of descent. It’s interesting to note that [participants’ crews] show a strong tendency to be on speed or fast on approach (Figure 1, p. 36). There is a steep drop-off on the slow side of the VREF [reference landing speed]. When we expand the view on the fast side, we find approaches flown well into the [VREF]+20 kt range.”
In their own standard operating procedures, participants typically call for flight crews to be at VREF at 50 ft above the runway threshold. But 2012 C-FOQA data showed that “the threshold[-crossing] speed distributions indicate that [participants’ crews] are carrying approach speeds to the threshold,” he said.
“Our [2012] flight operations events … are dominated by EGPWS [enhanced ground proximity warning system ‘GLIDESLOPE’ aural messages], which correlate with the high rates of descent indicated in the unstable approach events,” Charbonneau said. “A closer look at the rates of descent shows that [participants’ crews] are doing a really good job of achieving [the] 700 to 900 fpm rate of descent below 200 ft above ground [level]. However, there are still numerous C-FOQA data plots of 1,000 fpm rate of descent close to the ground — a clear indicator of a high energy condition.”
Most of the aircraft monitored by C-FOQA Centerline are certified with a 3-degree approach slope angle, that is, for touchdown points near the 1,000-ft (305-m) distance marker from the threshold. “Our average touchdown point in 2012 was about 1,700 ft [518 m] from the threshold, which is really not that bad,” he said. The data also revealed a “large number of landings” beyond 2,000 ft (607 m) and some beyond 3,000 ft (914 m) from the threshold.
One issue in steering committee discussion is an apparent trait of business aviation pilot culture: tailoring landing technique foremost to impress passengers. “It has been commonly accepted [for these pilots] to shallow landing flares to achieve a smooth or roll-on landing, thereby trading built-in safety margins for finesse,” Charbonneau said. “It really doesn’t make any sense.”
Runway length per se did not correlate with unstable approaches, but could be part of a combination of factors. “A strong indicator of runway excursions [risk] is a short runway while [the aircraft is] fast — [by C-FOQA definition, an] event that measured 80 kt of speed with 2,000 ft remaining,” Charbonneau said. “We’re beginning to see a trend (Figure 2) that when the runway is longer, pilots are willing to accept more [deviation from] precise landing performance … 2 percent of the flights … a big number.’’
Willingness to override the established criteria for stable approach is like one side of a coin, and reluctance to go around when unstable is the other side. “We can deduce … that we have a strong culture of not going around,” Charbonneau said (see “Inspiring the Decision to Go Around,” p. 28). “In the end it comes down to a culture of normalized deviation, or accepted noncompliance, or accepted nonperformance, or perhaps even a planned-continuation bias — regardless of what you call it, it is what it is. … This is exactly why programs such as C-FOQA Centerline are critical and needed more than ever.”
Some BASS attendees described their own mitigations. One practice is to avoid planned-continuation bias by conducting in-flight briefings for every approach and landing. The crew presumes a go-around will occur regardless of whether the flight is scheduled to conclude in visual or instrument meteorological conditions. They said this encourages vigilance, builds in a safety margin and reduces the trepidation and alarm that pilots may experience from lack of familiarity or recent go-around practice. “If we happen to see the runway, if we happen to be stable … then we’re going to land,” one attendee said.
Consistent With SMS
As an integral tool in the typical airline SMS, FOQA also supports today’s emphasis on predictive identification of risks, said J.R. Russell, chairman and CEO of Proactive Safety Systems and a United Airlines captain. “I’ll just say this about the predictive [method]: It’s quicker to the punch than proactive [or reactive methods],” Russell said. He described the methods in terms of practical issues in unstable approaches.
“[Imagine that] an unstable approach into airport XYZ has led to a runway excursion,” Russell said. “An investigation is done. Lessons are learned to prevent a similar incident. [The operator would] really want to be looking for latent failures that contributed to it. So that’s reactive safety.”
Imagine instead that, based on FOQA, the company had discovered an increasing rate of unstable approaches at XYZ. “Say they received a report from a pilot that said, ‘[We] were flying an approach at XYZ the other day, and the glideslope was [reported in notices to airmen (NOTAMs), because of construction, as] out of service. But because there were about three pages of NOTAMs for this particular airport, we missed it.’ [They flew] an unstable approach because they weren’t ready for [the non-precision approach] ahead of time. … [The operator could] put out a communication [to] raise everyone’s awareness. … That’s proactive.”
Imagine finally the operator working with the predictive mindset. “The company has the reports of unstable approaches at XYZ due to ‘glideslope inoperative’ because the runway was undergoing some type of construction,” he said. “[They also know] that airport ABC is about to have one of the runways go under construction. Do you really need to wait for crews to submit reports? Do you need to wait for your FOQA data to show a negative trend of unstable approaches?” So they also could use even wider communication as their mitigation.
Simulator Data Concepts
Recent experimental research for two U.S. military services concluded that both FOQA data and simulator operational quality assurance (SOQA) data could be integrated in novel ways relevant to business aviation, said Lou Németh, chief safety officer for CAE. The Air Force and Navy asked CAE, “Is there value in the simulator data and, more importantly, is there value in correlating what we see in the FOQA data [taken] off the airplane to the data that we see in the simulator?” A study concluded that this was “a successful demonstration. … The ability to automatically detect deviations from standard operating practice is a major innovation, and the objectivity was a major benefit.” In practical terms, it convinced him that data from a simulator can validate the efficacy of the entire flight training system.
Pilots’ exceedances of established flight parameters became measurable by collecting in advance the details of a specific operator’s SOPs on a standardized Microsoft Excel worksheet. “The SOQA algorithm looks at the Excel [worksheet,] then looks at the [simulator] data and compares it,” he said. The resulting report to the operator draws attention to exceedances/SOP deviations assessed as meaningful by the algorithm, how long they lasted and how they affected overall risk.
“The most frequent deviation from standard operating practice was a late extension of approach flaps [and] late gear extension,” he said. One bar graph, for example, showed that instead of nominal 1-g touchdowns (that is, one times the standard acceleration of gravity), some accelerations in the 5-g to 6-g range were reported. In training, this threat surprisingly could be overlooked. “First of all, you’re not going to feel that in the simulator because we can’t replicate this [acceleration] in a hexapod system for a sustained period of time,” Németh said. “So you rely on the data to tell you whether or not it was a good landing. But 5 g is twice the limit of the airplane. That airplane [actually would have] broken into several pieces on landing.”
Moreover, unloading inputs by a pilot to achieve an exceptionally soft touchdown — flagged by the algorithm as events with acceleration of less than 1 g — can be discussed with crews with the same objectives as Charbonneau’s warnings about the inherent yet underestimated risk of runway excursions.
Note
- The International Standard for Business Aircraft Operations (IS-BAO) is used by auditors trained and authorized by the International Business Aviation Council.