Confidence is stepping onto an aircraft. Why else would anyone board one? Just knowing that it works, it’s safe, and that the pilots up front can fly the thing – it’s something an estimated 0.1% of humanity take for granted every day. But before that 737 even makes its way out the factory door, who decides that its worth trusting? More importantly, how are those decisions made?
This week, AirlineGeeks was given the chance to meet up with test pilot and founder of Daedalus Aerospace, David Kern. Daedalus — recognizing a gap in test flight education — delivers training for aviation professionals, specifically through on-demand courses in flight test planning, flight execution, risk management and data analytics. Flight Test 101 is Daedalus’ introductory course, and this course dives deep into the principles that dictate a successful flight test program. Having been given access, we were provided with a new perspective on aircraft and system development.
We wanted to find out more about Daedalus, and more specifically, how these principles are applied in the real world. As Kern answered our questions, we gained a new appreciation of proper, planned flight tests. As Kern said, “if you don’t flight test your system — your customers will.”
Daedalus – What’s In a Name?
For the historically inclined reader, the story of Icarus is a familiar one. Flying too high and falling to his peril, it’s a story synonymous with the proud, the inexperienced, and the reckless — all of whom face the disaster of going “too far.” There is, however, Daedalus, who famously warned Icarus of the fate that would await him if he over-stepped the line. It’s this character that Daedalus Aerospace is named after.
AG: Many people are familiar with the story of Icarus, yet Daedalus – another character of the story – gets little attention. What makes his role significant? Could you connect that to the work you do?
DK: While I am not a student of ancient history, the sources I’ve read portray Daedalus as a far more interesting character than his son Icarus. I actually chose the name Daedalus Aerospace in his honor, with the idea that we can all strive to “Be a Daedalus, not an Icarus”. In Ancient Greek mythology, Daedalus was an inventor, craftsman, and artist. His works made him synonymous with skill and excellence. He is portrayed as a symbol of wisdom, knowledge, and power.
The most famous of the Daedalus legends describe him as exiled or imprisoned on an island with his son. In order to escape, Daedalus crafted wings for himself and his son Icarus from feathers and wax. Before their flight to freedom, Daedalus warned Icarus not to fly too high, because the heat of the sun would melt the wax. Icarus famously disregarded this advice, and even as their flight of escape from the island was successful, Icarus was tempted by hubris and flew too close to the sun. Although the legend says Daedalus escaped, Icarus plunged to his death.
Daedalus should be our model, because he was a craftsman that both knew how to effectively build the means of flight, and also had the skills to identify hazards and manage risks. I like to say that aerospace development is a team sport. We must be like Daedalus, able to not only creatively solve the challenges of flight but also apply risk management that allows confident progress without risk aversion.
AG: You’ve put together some impressive courses. What role do you aim to play in the flight test sector?
DK: I have always been fascinated with flight. Since becoming a test pilot, I’ve been fortunate to be a team member contributing to the development of some world-changing technologies — for example, I was a project pilot for the F-16 Automatic Ground Collision Avoidance System between 2009-2011. Along the way, I’ve had some incredible experiences and been trusted to fly amazing prototype systems. However, as my career progressed, I think I’ve realized that it’s far more gratifying to multiply these experiences for others by teaching and enabling them to develop the future of flight.
Today, we are living in an aerospace renaissance, with exciting projects and developments worldwide, at a far greater rate than we have seen for a generation. I want to educate and enable the engineers that are staffing up these aerospace teams for projects like drone autonomy, urban air mobility, supersonic/hypersonic aircraft, and commercial space launch.
Flight Test 101 – Principles for Test Programs
As we progressed through Daedalus’ introductory course, Flight Test 101, we came to appreciate that despite the public perception of flight tests (think Top Gun Maverick), the truth is somewhat less chaotic. In reality, flight test involves risk management and a well-planned, methodical approach based on two foundational concepts – that model validation and the so-called “build-up” approach.
AG: The Flight Test 101 course you deliver outlines significant principles that guide a test program, such as risk management and the “build up” approach to flight tests. Could you please explain how these principles are applied?
DK: Flight Test 101 describes the principles and concepts that guide modern professional flight tests. The two foundational concepts for flight tests are model validation and the build-up approach. Model validation is related to the concept of digital engineering or a digital twin. The planned aerospace system is defined in terms of mechanical structure, aerodynamics, electrical, thermal, data systems, and so on. This model of the system may be as simple as first principles, derived on the back of an envelope and roughly calculated to few significant figures – or may include computational fluid dynamics. The model is exercised to predict how the system is expected to respond in the planned flight conditions. These calculations are often generated before the physical prototype is fully manufactured, in the course of engineering iterations and design decisions.
Model predictions should include uncertainty factors for the calculated responses, to account for desired performance margins but also to identify potential exceedances or divergent responses. Flight test uses the modeled performance of the system to target test resources to critical conditions as well as to identify safety considerations. These test results are used to confirm the accuracy of the model or, alternatively, to re-tune the model to match the test results. By spot-checking at inflection points and the edges of the envelope, the development process can be confidently and safely accelerated toward operational fielding. Interpolation is sometimes acceptable, but extrapolation is generally not. Model validation is often described by flight testers using the mantra, “predict – test – validate”. The model predicts system behavior, the test evaluates the predictions, and the resulting data is used to validate and refine the model.
When choosing conditions for flight tests to address model validation, it is important to prioritize and sequence test points using the build-up approach. The term “build-up approach” is a simple way to express that initial flight test conditions should be identified with lower risk and higher confidence in the model, in order to progressively work out to the edge of the envelope. This is not a one-dimensional tactic but requires a collaborative and multi-disciplinary team to consider how to intelligently move into areas of lower model confidence, higher safety risk, higher energy states, lower terrain clearance, or reduced engineering margin.
The build-up approach sequences pre-requisite tests that contribute certainty through model validation. By starting with test conditions that have lower model uncertainty and lower risk, the refined model predictions can be fed back into subsequent tests to increase safety and efficiency. Using the model validation method and build-up approach, a flight test team may safely, efficiently, and effectively prove the capability of an aerospace system.
Flight Test and Commercial Aviation
AG: Flight test is often perceived as being closely tied to defense. How large is the commercial aviation sector in flight tests?
DK: Many people associate flight tests with military and defense, probably because of the high profile and public interest in government spending. However, the civil/commercial aviation sector also has a very significant scope of flight test work. I don’t have precise numbers for each segment of the flight test industry, and there is a certain amount of overlap because people often work on multiple projects, but I estimate that military and civil/commercial flight test sectors have similar employment numbers and economic impacts. Just think about the value of all the new airliners, or new business jets, or emerging multi-copter and drone projects! There is a lot of work going on across all these sectors.
AG: In the course, you teach that the goal of flight tests is to answer questions. What questions do you think need answering for future commercial aviation?
DK: The goal of a flight test is always to answer questions, while also managing risk to an acceptable level. These questions can be quite straightforward, such as “how fast will it fly?”, or “what is the takeoff distance at the maximum weight on a standard day at sea level?”.
Those are simple questions, although answering them may require some risk management considerations. However, some of the more difficult and important questions that flight test answers are things like “is this aircraft design compliant with the applicable civil certification standards?”, or “does this weapon system fulfill the military specifications and operational utility requirements?” Those questions generally can’t be answered in a single flight, and usually require quite a bit of specialized technical documentation.
Future commercial aviation will have to answer all these types of questions but in different contexts. Important questions will include things like “how well does the fly-by-wire system for this multi-rotor air mobility system handle the urban wind jungle, and what limits should be documented to protect the operators/passengers?”, or “what combinations of supersonic Mach and altitude result in an acceptable noise signature for a flight over populated areas?”
Learning Lessons for the Future
In 2019, three letters changed the face of commercial aviation – MAX. Almost as if Boeing flew too high, the tragedy and fallout of the 737 MAX reminded the world that although a system might work well in theory, proper test flight would have taken much more into consideration. We asked Kern for his thoughts on the 737 MAX, as well as what lessons are needed as aviation experiences a new renaissance in the 21st century.
AG: I’d like to get your perspective on the Boeing 737 Max. While MCAS is not a unique technology (I can think of character augmentation systems used by Empires Test Pilot School, for example), do you think it received adequate testing? Is this a textbook lesson in poor planning?
DK: Much has been written about the Boeing 737 MAX, and I think the best reference is always the
Joint Authorities Technical Review . The takeaways that I like to emphasize are 1) the necessity of thinking holistically about aerospace designs from a systems engineering perspective, and 2) recognizing the complex nature of human-machine interfaces. We need to design future aerospace systems (and really any safety-critical system) not just so that they CAN work correctly — but also to account for failure modes, human error/reaction time, and engineering margins.
AG: In the course, we learn that a successful test doesn’t necessarily mean we get the results we would like. How can we create a team environment where engineers and professionals can feel safe enough to speak up and re-examine models and test again?
DK: One of the most important concepts for aerospace development is that a successful test simply means that you set up and executed a controlled experiment that resulted in meaningful data. It doesn’t mean that your product functioned particularly well. You might actually be disappointed with the results from a business sense or product development perspective — a successful test might send you “back to the drawing board”, but the test is successful if you set it up to answer a meaningful question and got data that supports a decision.
This is why integrity and a safety culture are so important to engineering in general and flight test in particular. Sometimes, the person that knows the most about the design of the system under test, is the junior engineer (or even the intern!) in the back of the room. That person needs to be empowered to speak up to prevent unsafe outcomes, both for the test crew during development and also for the eventual customer. There are many ways to develop this safety culture, but all of them take ongoing effort at two-way communication between the engineering and test teams, plus management commitment to integrity and safety as the highest priorities.
AG: Computational fluid dynamics have improved over the years. Do you see an increasing reliance on CFDs and less on test pilots in the future?
DK: There are many areas of aerospace development where our tools have significantly improved, including CFD and finite element analysis (FEA). In a more broad sense, these tools are sometimes called digital engineering or using a digital twin. While those tools are certainly needed to create future designs and accelerate the engineering process, I know that flight tests will always be necessary to validate the models.
Remember that one of the fundamental principles of flight tests is Model Validation. Models are dependent on assumptions, which in a worst-case scenario leads to garbage in – garbage out… but can also have the undesirable effect of a point-condition optimized design. The only way to really prove an aerospace system is with an actual flight test.
Remember, if you don’t flight test your system — your customers will. You don’t want potential problems to be discovered “in the wild”, without any risk management processes in place or experienced flight testers to identify the issues before they escalate and become an incident/accident. From another perspective, an overreliance on digital tools can result in long, expensive delays with endless digital iteration and optimization. It’s easy to forget the effects of real-world manufacturing tolerances under operational flight conditions when you’re staring at a CAD drawing. Early flight test (with some risk-management controls in place) can be an engineering development accelerator, by gaining confidence in the reliable areas of the model and quickly showing where our assumptions were flawed.
From Start to Finish
Having gleaned much from Kern and Daedalus Aerospace, we were left with a deep appreciation for the entire lifecycle of an aircraft. Whether we are boarding a short, regional flight, or relying on emergency relief from above, flight test is the behind-the-scenes magic that ensures confidence in any flying machine.
From the very start and right through to the finish, the flight test provides the right answers to the right questions. Maybe that’s why the test pilot has the right stuff. And from what we gained from Daedalus, this is one ingredient that will never grow old.
