It has long been a tenet of flight testing that you never risk more lives in a test program than necessary. This has been true in both the professional civilian and military worlds for a long time, and it is codified in the Experimental/Amateur-Built world via our operations limitations. Section 10 states that “During the flight-testing phase, no person may be carried in this aircraft during flight unless that person is essential to the purpose of the flight.”
The single seat, one-off KK-1 couldn’t accommodate a second pilot—you had to fly it solo the first time.
Up until recently, the FAA interpretation of “essential personnel” has meant that, for single-engine Experimentals, only one person can be onboard. It has been argued by many that an additional person is useful to take data, look for traffic, and operate systems—but that is not the interpreatation that counts. The FAA has maintained that if the pilot can’t do those things in a simple aircraft, then they aren’t truly qualified to be the one doing the flying. Basic risk management says that the fewer warm bodies you put in harm’s way, the better. This has been true since the E/A-B world was created—until now. With the issuance of a new advisory circular, the FAA is going to start allowing the option of having an additional pilot in the cockpit of two-seat (or more) Experimental aircraft during Phase 1 testing. By so doing, the FAA believes that they can reduce fatal accidents during the critical early hours of a new Experimental aircraft’s operating life.
How did we get here? It all started with a letter from the NTSB to the FAA issued back in 2012 giving the agency specific suggestions of ways to reduce the accident rate in Experimental aviation. When compared to the certified GA world, Experimental aviation looked particularly bad, and the NTSB had studied the accident reports and records to provide a “top ten” list of improvements it would like to see—and that the FAA could implement. While not binding on the FAA, the FAA nevertheless was obliged to look at the areas suggested by the NTSB report and either implement the suggestions or provide rationale for why they were not going to do so. Statistically, the two greatest causes of Phase 1 accidents were fuel system problems (leading to engine failures) and loss of control accidents—and the two are not mutually exclusive.
This figure from AC 90-APP is a stark reminder that most accidents in Phase 1 flight testing occur in the first few hours of testing—and 18% happen on the first flight.
The argument for two in the cockpit goes like this: A builder has finished their new airplane, but they are not current, not experienced in type, or not experienced with flight testing to feel comfortable doing the first flight on their own. Rather than simply relinquishing the airplane to a qualified friend or other pilot, they stubbornly maintain that they are going to be in the cockpit for the first flight of the plane they built—and since the rules say they have to do it alone—they do. That lack of experience or currency (or both) comes to bite them when the engine stumbles, or the airplane is more than they were expecting, and the subsequent loss of control or untoward arrival to earth ends badly. Now, it is argued, if they had a qualified safety pilot in the cockpit with them, then when the engine stumbles, they have someone to take over and save the day. That makes sense right up to the point when you ask, “So what is the inexperienced pilot doing there in the cockpit anyway?” And an unemotional risk analysis will support that point of view—there is no reason.
But human beings are far from unemotional, and often far from logical. Imagine that you are a flight advisor who has been asked to help a person plan their first flight. The builder/pilot maintains that the airplane will not fly without them in the cockpit—and they want an experienced pilot to come along. You advise them of the facts, but they maintain that if they don’t have a qualified pilot to ride along, they’re going to do it solo. Now that is a dilemma if your goal (as an FA) is to try to keep things as safe as possible—and save lives.
So what if it was legal to have a qualified pilot aboard? Is this really a good idea? The truth is that if there are, say, ten fatal accidents a year on first flights (it is more or less the case, statistics being what they are), half of those are engine failures, half are loss of control. If an inexperienced pilot is alone in the airplane, let’s assume that all are fatal. But if there is an experienced pilot onboard, a percentage may not be—because the experienced pilot performs a successful forced landing, and knows how the airplane handles and stalls. You will probably cut the number of fatal accidents more than in half. Sure—in some remaining percentage of cases, you still have a fatal accident—and now you kill two people instead of one. The question is, by putting the experienced pilot in the cockpit, do you cut the overall number of fatalities to a lower number than before? The FAA has numbers and accident statistics that show that they believe that you will. By allowing the extra person, they acknowledge that you will still kill some people—but it is a lower number than if you forced all the flights to be solo.
This is a hard reality for experienced risk managers to swallow—this author included. Having been in the test business for over forty years, I always try and cut down the exposure to the fewest number of souls. But in those cases, I am always dealing with qualified individuals who are going to follow the rules because that is their job. The Experimental/Amateur-Built world is different, and we have to recognize that. We do not have the force of employment rules to keep pilots on the straight and narrow—all we really have is peer pressure. The process allowed by the FAA’s new advisory circular is not as simple as just finding a pilot to sit in the other seat—there is a lot that goes into determining who is qualified. But that is simply an indication of just how seriously the FAA takes the risk management process—if there are going to be two people in the cockpit, then they want the additional body to be valuable to decreasing the overall risk.
First things First
There is no getting around your operating limitations—they are part of your airworthiness certificate, and they have the force of law. And every set of ops lims written to this point (that meet the guidelines of FAA headquarters) state that only essential personnel may be aboard during Phase 1. That means a single person for all reasonable modern homebuilts. So the first thing that has to change in order for the dual-pilot option to be in play is—the ops lims. A provision of the approval of the new process will be that DARs and FAA inspectors will issue an operating limitation that allows a second person on board, so long as they meet the requirements of the advisory circular as a qualified pilot. Sorry—if you have a set of ops lims issued before the approval of the AC, you’ll have to get them modified to be strictly legal—and the AC specifies how to do this.
There are some general rules that must be satisfied in order for a builder to consider the second pilot option. The builder/pilot must own all or some portion of the aircraft and hold a valid pilot license appropriate to the aircraft. They must have a current flight review, and meet the recent flight experience requirements in order to carry passengers. And they need to list the name of the additional pilot in their own logbook for each flight in which the additional pilot is carried. Aside from the ownership requirements, these are easily met to be legal for flying the airplane anyway, and add no further burden.
As important as having the proper operating limitations is having the right aircraft. Not all Experimentals will qualify to be eligible for the new program of dual pilots during Phase 1. There are three requirements listed in the AC—and all must be met. First, this must be an Experimental/Amateur-Built or Experimental Light Sport Aircraft. Second, it must be from a kit—no plansbuilt aircraft are allowed (this is to assure a certain level of conformity to a standard design). Third, the aircraft must have fully-functioning dual flight and power controls—no stick-only backseat aircraft need apply.
The purpose behind all three of these rules is to ensure that the aircraft is a well-known “standard” type that has known flying characteristics. We all know that just because it is a kit, and just because it has flown before, that doesn’t ensure that it flies well—but at least the pitfalls of the design should be known. If an aircraft is built from a kit, but has significant modifications, it might not be a good candidate for the program. While the AC doesn’t specifically address airframe modifications, the most common mod that builders make is to install a different powerplant, and because powerplant failures account for one-third of all E/A-B accidents, the AC goes on to have specific requirements for the powerplant as well.
Even the ViperJet—complex by anyone’s definition—requires only one pilot to fly. Interestingly enough, it would not qualify for the dual-pilot program because it has a turbine engine.
In order to qualify for the Additional Pilot Program, the aircraft must be equipped with an engine recommended, supported, or provided by the kit manufacturer. Accessory modifications (such as after-market fuel injection and electronic ignitions) are allowed. Unless the manufacturer directly recommends auto conversions, they are not allowed. Turbine engine aircraft are excluded from the program, but this won’t affect too many builders. The bottom line is that if you are doing something radically different than what has been done before, the airplane has a significantly higher risk and therefore should be flown solo.
In addition to having an eligible engine, the powerplant must be tested according to a long-standing flight-testing advisory circular—AC 90-89. The testing must be documented and signed off in the aircraft logbook in order to qualify for the Additional Pilot Program (this can be done by the builder). Tests required include:
• Mixture and idle speed check
• Magneto check
• Cold cylinder check
• Carburetor check
• Fuel flow check
• Unusable fuel check
• Compression check
While this might, at first, seem to be a long list, most of these items are done by a conscientious builder before the first flight today—regardless of how many souls will be onboard. Such tests are a great way to get to know the airplane, provide numbers on fuel flow and tank calibration for a POH, and familiarize the pilot with engine starting procedures, operating characteristics, and normal instrument readings.
Test plans are divided into Initial Tests, which require a very qualified additional pilot (or a solo flight), and tests which will fill out the Phase 1 program and require less of the additional pilot.
The FAA recognizes that there is a difference between initial testing of an aircraft for stability and control, and subsequent performance testing and opening up of the envelope. In order to mold this into a Phase 1 program, they have tables in the advisory circular that show typical testing required for initial testing and follow-on tests—and they use these tables to define the skill-set for the additional pilot based on where the aircraft is in Phase 1. Simply put, an additional pilot needs to have a higher level of qualifications for initial flight than for subsequent flights where the aircraft has begun to prove itself. To operate as the additional pilot during the early phases of the test program, a pilot must meet a set of criteria delineated in the AC, and the additional pilot must continue to meet these criteria until the aircraft (and the builder/pilot) has met the initial testing requirements. These requirements include eight total hours of flight time and specific tests such as ground runs, taxi tests, first flight, pitot/static checks, wings-level stalls, and approaches to accelerated stalls.
Once these initial tests are complete, logbook entries must be made attesting to these results, and the requirements for the additional pilot drop significantly. What this really does in a practical way is to allow a highly qualified pilot to be present in the cockpit during the initial, more risky phase of Phase 1, and once that period has passed, Phase 1 will begin to allow such activities as pilot transition training by CFIs or other pilots experienced in the aircraft. It also allows a less-experienced additional pilot to go along to take notes and gather data if the builder/pilot feels that is necessary. This helps to satisfy the demand by those who want to get trained in the aircraft they built without having to wait for Phase 2 to do so.
You don’t have to have the last name of Yeager, Crossfield, or Rutan to meet the standards of a qualified pilot for the initial flight phases—but the process delineated in the advisory circular will take a close look at a pilot’s recent experience, experience in type, and overall flight test experience before blessing them to occupy that second seat on a first (or subsequent) test flight. You don’t have to be the greatest pilot the world has ever seen—but experience in the realm of flight testing is going to be required to meet the initial level of the bar.
To qualify as an additional pilot during initial flight testing, these recency-of-experience criteria must be met.
Rather than defining a single set of mandatory requirements to determine a pilot’s eligibility to act as a qualified pilot, the AC provides two scoring matrices—one to determine a pilot’s recent experience, and one to look at their overall accumulated experience to determine if they can achieve enough points to qualify. For instance, in the recent experience category, they get points for takeoffs and landings in the previous 90 days, time in type, and total hours of flight time. For each of these categories, there are absolute minimums—not meeting them kicks the pilot out of the matrix, and out of the program. Not having at least ten takeoffs and landings in the previous 90 days, having no time in type, or having less than 500 hours will disqualify a pilot—among other things.
If a pilot scores high enough on the recent experience matrix, they move on to the experience matrix—a measure of their overall experience in the type of aircraft and test flying in general. Points are given for time in the same category and class, the same type of aircraft, and time in the same model. On the flight test side, points are earned for previous Phase 1 experience, for time in the same configuration of vehicle (canard or high performance), for first-flight experience, and for the level of pilot certificate held. Graduates of a qualified test pilot school are given a very large number of bonus points—as you’d expect—but test pilot school is certainly not required to qualify.
A passing score on both the recent experience and overall experience matrices allows a pilot to act as the additional pilot—but does not require them to do so. That should be remembered. It is still a personal and professional choice to act as an additional pilot, and builders need to recognize that many pilots may decline based on risk minimization analyses. This is, after all, a voluntary program for both the builder/pilot and the additional pilots.
Experience counts for an additional pilot, and this matrix will help to determine if they have what it takes.
So how is all of this supposed to be enforced? Is the FAA going to check logbooks and score sheets before every flight? Will there be surprise inspections at dawn on first-flight days? We haven’t seen a plan for this part of the program, but judging by how the FAA checks medicals, I would suggest that this process will be enforced by peer pressure and the honor system. Let’s be honest—we all know of people that are doing Phase 1 with more than one person, despite what their op lims say. We also all probably suspect that old Fred down in the end hangar hasn’t seen a doctor since before the days of computerized medical forms. Virtually all of general and Experimental aviation is enforced by the honor system—the occasional ramp check notwithstanding. Compliance is assumed until an incident or accident takes place. And, unfortunately, E/A-B airplanes in Phase 1 have way more than their share of incidents—so the odds of being found out in the case of a problem is probably greater than if the average GA pilot flies the wrong altitude while VFR on an airway.
We expect that there will be many builders who are overjoyed at the prospect of being able to find a qualified pilot to accompany them on their first flight. We expect that there will be a great many qualified pilots who will not do so. We expect that there will be pilots who qualify under the program who will be happy to oblige, or who agree to go along because they are afraid of what will happen to a rookie if they don’t. My expectation is that many of those who are best qualified will look at all aspects of risk for a particular airplane, pilot, and situation, and use their best judgment to fly, not fly, or simply intensify their counseling of the pilot one way or another.
Many will read the title of the new advisory circular, see that their ops lims allow two people, and grab old Fred down at the end of the row to go along and “keep them safe” on their first flight—but that is not the intent of the authors of the new advisory circular. The FAA authors understand that they won’t eliminate all of the risks of Phase 1, and that pilots die the way it is now. Under this new program, some will still die, but they believe that the statistics will show that fewer overall will come to a bad end. And in the ruthless world of risk management, a reduction is always a win.