FSF demo brings airline safety system to bizav

 - August 29, 2007, 12:09 PM

Flight operational quality assurance (FOQA) programs are finally seeing application in business aviation, and in early February the Altria flight department revealed that it has been working with the Flight Safety Foundation in developing a FOQA demonstration program. FOQA compares data downloaded from an aircraft’s digital data bus with aircraft limitations and operator standards, according to Jeffrey Sands, director of flight operations at Altria Corporate Services. The data helps the operator identify safety or operational issues and modify policies, procedures and training to reduce risk.

Airlines have been using FOQA since the 1990s, after resolving issues about data sharing with pilots’ unions, and the process has saved money and improved safety.
Robert Vandel, FSF executive vice president, and the Flight Safety Foundation in 1992 conducted a study that recommended the creation of a flight data monitoring system for airline risk management. The term “flight data monitoring” was soon replaced by FOQA. “Pilots didn’t like the idea of monitoring,” Vandel said.

Later in the 1990s, Mobil flight department manager Pat Andrews suggested to Vandel that it was time for business aviation to adopt FOQA, and Sands began looking into the subject. “I soon discovered it was going to be a complex undertaking and probably outside our capabilities as the single operator of several aircraft,” Sands said.

In 2000, Sands discovered that the FSF was working on a FOQA demonstration program. A member and later chairman of the NBAA Safety Committee, Sands encouraged NBAA to partner with the FSF on bringing FOQA to business aviation. The FSF created a task force to study FOQA for business aviation and how it differs from airline FOQA.

At its heart, FOQA is a tool that helps an operator identify risk. It complements safety management systems such as the International Standard for Business Aircraft Operation (IS-BAO). A safety program such as IS-BAO, however, requires identification of risks, management of those risks and measurement of whether the risks are being minimized properly. “I hold that identifying risk is the hardest part,” said Sands.

The two types of risk, he explained, are known and unknown. “The only thing we can do anything about is known risks,” he said. “We have no idea how big this pool of unknown risks is. That’s one of the areas where FOQA can help.”

In trying to measure risk, business aviation has had to rely on statistics from airline operations, because there isn’t much data on business jet operations; with few exceptions, most business jets belong to single-aircraft operators or operators of small fleets. “Within corporate aviation and to some extent Part 135,” Sands said, “we have fewer accidents and in most cases [they] are not necessarily investigated with as much vigor as a major airline accident. We’re left with very little data and information upon which to assess our own risk.”

For any safety program to be effective, the operator needs accident and incident data from similar types of operations, adjusted for the unique characteristics of that operation. “You can brainstorm and get some of it,” Sands said, “but when you measure the way that FOQA does, you’re presented with information that is compelling.” This is because FOQA delivers summarized information covering a specific period of time, he said, “and it just hits you right in the face rather than looking at issues on an individual basis. You can see relationships better, and you’re not chasing individual issues. Some might be trends, others are not.”

Approach and landing accidents are a primary focus of the FSF demo because they continue to occur fairly frequently, accounting for 50 percent of accidents during any five-year period since World War II. “We know that approach and landing accidents are caused in a majority of cases by unstabilized approaches,” Sands said, “or they are a contributing factor. Do you have a real concept of how that measures up in your own operation? If we can stamp out those, we could take out a big chunk of risk.”

The FSF recommends stabilized approach criteria, but how is the operator to know whether pilots are adhering to these criteria? With FOQA, operators can measure aircraft performance against those criteria.

Sands was careful to address any potential legal issues before embarking on the FSF FOQA demo. The biggest issues were protection of the data, which should belong to the operator, and protection of the pilots. “That was a great concern for us,” he said, “because we knew that we needed the [support] of the pilots to make it work.” Corporate operators generally don’t have unionized pilot workforces, so pilots might not enjoy the same protections as their airline counterparts. For airline FOQA, the unions are signatory to an agreement between the airline and the union to protect the pilots, and the union helps analyze data and investigate exceedances.

For the FSF demo, the Foundation acts as program administrator, and participants must sign an agreement with the FSF. This document includes provisions that protect pilots, such as agreeing not to use data for punitive reasons, Sands explained.

Another key element of business aviation FOQA is that it shouldn’t be used to identify an individual pilot’s performance or help spotlight a “rogue” pilot who flies dangerously outside normal limits. The majority of pilots, he said, “are professionals and want to operate the aircraft professionally. If you don’t know about a ‘cowboy’ [in your flight department], you’ve got other problems.”

To participate in the FOQA demo, Altria had to add a quick-access recorder (QAR) to two of its jets. This works best with later- generation airplanes that have digital flight data recorders (FDRs). The QAR plugs into the FDR data bus and records all the parameters the FDR records, at a rate of once per second. Altria bought Avionica QARs for $5,495 each and installed them under an STC at a cost of about $2,000 per airplane (including the STC price).

To get the data from the QARs, Altria had to buy a download kit, which also cost $5,495 and includes a special cable and a software license. This kit can be used to service all the QARs in a fleet.

Once the data is downloaded into a computer, the data analysis–the most expensive part of the FOQA operation–takes place. The raw QAR data is useless and must be run through analysis software to extract useful information. Altria sends the data, which is compressed and encrypted by the download software, via secure Internet link to an analysis vendor, in this case Austin Digital of Austin, Texas.

Austin Digital charges $10,000 per airplane per year for its analysis services, but Sands and the FSF believe that should drop as more operators elect to participate in the FOQA program.

When it receives the data, Austin Digital de-identifies it, removing anything–such as date information–that could show who was flying a particular flight. “We don’t want to have a means to identify a particular pilot,” Sands said. When the analyzed data returns to Altria, the company safety manager reviews it to make sure there is no identifying data. Then the company can begin to use the data that has been extracted.

Austin Digital supplies the data in chart and graph format. The operator has to work with the data vendor to set parameters for each area of interest. For example, in the stabilized approach, the operator can set parameters for airspeed at a certain height above touchdown; Altria used 500 feet in the demo. “We know that if there’s a tendency to carry excessive speed, there is an increased risk of overrun,” Sands said. “So what we would receive is a raw number or percentage number of approaches that were triggering as ‘fast’ at some point throughout final approach.”

Other criteria for stabilized approaches might include, for example, one dot left or right of the localizer, on speed, with a stabilized power setting. “Every one of those things we can measure with FOQA,” Sands said. “And from [a particular] gate down to touchdown you need to be within those specified criteria for it to be considered a stabilized approach. Excursions outside those criteria are grounds for a missed approach.”

What does an operator do with this data? The FOQA data helps identify what the big issues are so the operator can apply its resources to those rather than waste money on areas that aren’t going to deliver a safety or cost benefit. And the data helps the operator identify a risk and how often the airplanes are subjected to it.

In terms of using the data to modify pilot behavior, the data could be used to communicate to pilots that certain operations involve more risk than the flight department is willing to accept. “[Pilots are] professional and competitive by nature,” Sands said. “I think a good strategy is reviewing the results. If [the pilots are] aware they’re being looked at, it goes to reinforcing what’s important. If you stress the approach criteria, and you’re showing results and where there might be an opportunity to improve, you’re going to see some improvement.”

This works only to a certain extent, Sands cautioned, before an operator might have to use the data to modify the company’s training program; but that is another benefit of FOQA. “There is an obvious linkup between FOQA and training,” he said.

Is it hard selling FOQA to corporate pilots? Not according to Sands. “If you focus on FOQA as a tool to measure the overall operation and less [on] the individual information that is being recorded, you probably stand a better chance of selling the concept.”

And current tools, other than FOQA, aren’t quite doing the job. “As corporate operators,” he said, “we have sophisticated standard operating procedures, we have stabilized approach criteria, and we train to them. But something just tells me that we haven’t eradicated all risk. We know it’s out there, we put out these defenses, but we know something [could happen] under the wrong circumstances. So I think it’s time to turn to new, more sophisticated tools if we want to improve our already excellent safety record.”

Then, too, there is the phenomenon of how people behave when they know they are being observed. A cargo Learjet pilot who decided to roll his company’s Learjet 35 as a way of celebrating his last cargo flight on January 10 might not have done that if he knew that the stunt would show up in a FOQA analysis.

Another facet of the observation phenomenon is that there is a higher rate of accidents in aircraft flying empty legs, indicating that sometimes pilots fly differently when there are no passengers. “With FOQA, you always have a passenger,” said Jim Burin, FSF director of technical programs.

“Our vision,” said FSF’s Vandel, “is to get enough operators [using FOQA].” He wants to see the results of fleet-wide FOQA programs available for all participants to view, to help identify geographic and other safety trends. “We’re getting smarter as we get more information on this,” he said.