NTSB’s Michael Graham Weighs In on Key Safety Issues

Q: NBAA recently received an urgent recommendation from the NTSB asking the association to share the importance of using properly trained pilots to perform post-maintenance stall testing. Could you share more about why it is so important that operators utilize experienced pilots for these tests?

Post-maintenance stall tests demand a level of pilot expertise that goes beyond what is required for normal line operations, and the NTSB’s recent investigations into two fatal Hawker aircraft accidents make this point unmistakably clear. In both our 2024 Hawker 900XP and 2025 Hawker 800XP investigations, we found the flight crews were fully qualified to operate their aircraft, yet their training, experience and procedures available to them left them unprepared for unexpected stall behaviors that emerged during post-maintenance test flights.

These tests present unique hazards to operators given the highly sensitive nature of even minor imperfections on the wings of these aircraft, such as discontinuities or ridges in the sealant between the leading edge and upper wing skin, light ice accretion that may be visually imperceptible, or other wing component condition defects. Such anomalies can cause the aircraft to stall before the stick shaker or stick pusher activates, leaving pilots with no warning that a stall is imminent. Without proper experience or training, flight crews might be unable to effectively respond to these unexpected, unfamiliar conditions and recover the aircraft.

Experienced stall-test pilots understand these aircraft-specific sensitivities, have practiced recovery from fully developed stalls and unusual attitudes, and know how to manage the aircraft when it behaves outside the envelope. As we stated in our report, it is not safe to assume that any pilot is qualified for these post-maintenance stall test flights without specific, specialized training.

Post-maintenance stall test flights introduce increased risk, compared to normal operations, that needs to be mitigated accordingly. Using properly trained and experienced stall-test pilots is essential to ensuring these flights are conducted safely.

Q: Loss of control inflight, runway incursions and excursions and controlled flight into terrain are always important safety areas for business aviation operators to focus on. Given the NTSB's investigative experience with these kinds of accidents and your background in business aviation, can you share a few general observations about these safety areas and potential broad steps the industry and government could take to address them?

All of these issues are stubbornly persistent in accidents we continue to investigate, which is especially frustrating considering how many tools are available to help mitigate many of these circumstances.

For loss of control inflight, the NTSB for years has asked the FAA to require Part 135 operators to incorporate upset recovery training, checklists and procedures into their training programs. This stemmed from the 2007 Marlin Air crash outside of Milwaukee, claiming the lives of all six individuals on board, which our investigation determined likely resulted from the pilots’ mismanagement of an abnormal flight control situation through improper actions, including failing to control airspeed and prioritize control of the airplane.

We found that pilots would benefit from training and readily accessible guidance indicating that, when confronted with abnormal flight control forces, they should prioritize airplane control (airspeed, attitude and configuration) before attempting to identify and eliminate the cause of the flight control problem. As a pilot who has gone through upset recovery training myself, I know how valuable it is to experience these events in a safe environment to learn how to properly respond to them. Operators should take the initiative and ensure initial and recurrent upset recovery training is provided to their flight crews.

Runway excursions remain one of the leading causes of general aviation accidents, and one of the most common contributing factors we see for these accidents is the failure to establish and maintain a stabilized approach. This failure, or continuing an unstabilized approach, could lead to landing too fast or too far down the runway, potentially resulting in a runway excursion, loss of control, or collision with terrain. Regardless of aircraft type, pilot experience, or whether operating under IFR or VFR, a stabilized approach is essential for maintaining control and ensuring a safe landing.

To achieve this, pilots should follow standard operating procedures and industry best practices for stabilized approach criteria. These include maintaining a normal glidepath, proper airspeed and descent rate, correct landing configuration, appropriate power settings, completion of landing checklists, and stable alignment with the runway requiring only minor corrections. In most cases, the approach should be stabilized by 1,000 feet in instrument conditions or 500 feet in visual conditions. If the approach becomes unstabilized at any time after these points, a go-around should be executed immediately.

Regular practice of go-arounds and missed approaches, along with preflight planning for these scenarios, help ensure pilots are prepared to respond effectively to these unexpected conditions. Additionally, and perhaps most importantly, pilots must not allow operational pressure, continuation bias or last-minute changes to influence their decision-making. Having flown under demanding conditions, I know those pressures can be real, but pilots should never attempt to salvage an unstabilized approach.

Finally, many of the accidents we investigate involving controlled flight into terrain occur during nighttime visual conditions. In many of these cases, flight crews are in communication with air traffic control and receiving radar services, with neither pilots nor controllers aware that the aircraft is in danger. We have found that increased altitude awareness, better preflight planning, following IFR practices such as climbing on a known safe course until well above surrounding terrain, using sectional charts or other topographic references to increase familiarization with the terrain, and incorporating GPS-based technologies like enhanced ground proximity warning systems can all help avoid these accidents.

Q: With the expanded implementation of the FAA’s 14 CFR Part 5 Safety Management Systems (SMS) well underway, what should operators who are covered by the mandate keep in mind as they work to implement, document and maintain a Part 5 compliant SMS by May 2027?

The most important step operators can take as they get ready for SMS compliance is the very first step – ensuring leadership is not only on board with the concept but completely bought-in with a full understanding of their role and responsibility as an accountable executive. If company leaders are not willing to be the accountable executive, or if they simply sign a piece of paper, it undermines the entire concept of an SMS. It’s important to understand that leadership buy-in does not mean they are writing a blank check for safety, which is a common misconception.

To the contrary, it means the C-suite is putting a system into action that will ultimately save the company. People may roll their eyes when SMS is discussed as black ink on the budget sheet, but when calculating the cost of even one accident, the benefits of SMS in both human and monetary capital become clear. The accountable executive within the organization needs to not just understand their role within the SMS structure, but take action to ensure that the goals and objectives of the SMS are met.

Additionally, as the compliance date approaches, operators should be thinking about their confidential reporting system. In order for an SMS to work, every employee in the company needs to be engaged and understand their roles and responsibilities.

In an SMS, each employee is the safety officer. This is because each employee has defined roles and responsibilities when it comes to safety. Employees should not only be able to articulate their roles and responsibilities within the SMS, but they must have a confidential system to report risks or hazards that they believe may lead to an incident or accident. The confidentiality of the system is critical, but just as important is ensuring there is a feedback loop. If employees use the confidential reporting system but reported hazards are not mitigated and there is no communication to employees about the issue, it will discourage that employee – and likely other employees – from utilizing the system in the future.

Q: Overseeing a safety management system is obviously a critical task in any operation, as you know from your previous experience as an SMS manager and more current NTSB accident investigations. What ideas can you share to help SMS managers strengthen their SMS and ensure its effectiveness?

An SMS should never be static by design – there is always new data to monitor, policies to evaluate and risk mitigations to put in place. I’ve seen and heard some managers view SMS as checking a box. That’s not what safety is about, and it is certainly not proactive safety management.

Additionally, as I mentioned above, a fully implemented SMS necessitates buy-in from the top down and organizationwide. SMS managers need to be visible to all departments within the organization, from the C-suite to the IT department to human resources and everything in between. Every individual and department within an organization has responsibilities and duties within the SMS structure, and SMS managers need to stay visible and ensure that message is being carried out. By being accessible, employees are more likely to not only remember their responsibilities within the SMS, but they’re also more likely to ask questions, report hazards and suggest improvements if they have experienced issues within the SMS structure – all things that I would say are indicative of a positive safety culture. If employees don’t know where or how to report those issues with the SMS itself, trust in the system will erode and it will not be successful. SMS managers are responsible for ensuring everyone from the accountable executive on down is cultivating a positive safety culture.

Q: What role do you and the NTSB see different technologies playing in enhancing safety today and in the future?

Technology has come a long way since I started flying, and it has without question improved safety by leaps and bounds across the industry. I see two categories of technological advancement in aviation safety, and each is critical to flight operations.

The first category is technology that improves flight operations, everything from glass cockpits to ground proximity warning systems to ADS-B. Many of these technologies were originally designed for greater efficiency, but ultimately many of them have helped the industry achieve the sterling safety record we see today. Many of these technologies introduce an additional redundant system, adding another layer to the Swiss cheese model of aviation safety. Of course, as with any new technology, the key is how they are implemented and integrated. In my view, this will continue to be a critical question for the industry moving forward. Technology on its own does not improve safety; it improves safety when humans are properly trained to understand and interact with it. Human factors system safety assessments must be a part of any new technologies impacting flight operations to ensure we understand how humans react to the new technology before them.

The second category is technology that allows operators to analyze their data. Coming from an operator, it always amazed me how much data we had access to. Today’s technology allows for automatic downloads, some even in real time, from every flight. But, just like technology alone does not necessarily improve safety, data is only useful when analyzed properly. As the industry discusses potential uses for artificial intelligence, I see data monitoring as a key area where AI can be incorporated so that operators can maximize their data, discover problematic trends early and ensure risk mitigations are working as intended.