Keynote remarks delivered at the American Bar Association
Air and Space Law Committee

I want to begin by thanking the American Bar Association  for giving me the opportunity to address this group today.

I’ve been involved in aviation safety for a long time, going back to my days as a naval aviator serving as the squadron training and safety officer, to the National Transportation Safety Board, the Federal Aviation Administration, and tangentially, in my role today at Rockwell Collins.

The last time I talked with an audience of attorneys like this, I talked about runway safety, which is a particular passion of mine.  I discussed the historical trends, talked about the Air Traffic Control and pilot training being conducted, the new lighting and signage requirements, the emphasis on solving airport “hotspots”, and the new technology being rolled out at the time, such as Airport Surface Detection Equipment – Model X and Runway Status Warning Lights.

Then, I played a video (below) that showed a commercial transport crew landing their aircraft at an airport with parallel runways.  They were landing on the center runway and were told twice to hold short of the inboard runway after rolling out.  This was due to a departing aircraft.  The commercial transport crew acknowledged the transmission both times and yet, upon exiting the high speed taxiway, they continued onto the inboard runway, where the departing aircraft overflew them by a couple of hundred feet or so.

And my plea to the group was, despite all the training, signage, lighting, and positive two-way communication, the human in the loop makes mistakes.  And I asked the group to help the FAA solve these types of issues.

My own view is they can only be solved with automation, including positive aircraft control where needed.  For those of you familiar with it, this was the vision, in part, behind Nextgen, originally the Next Generation Air Transportation System, NGATS.

That is, to get to the next level of safety and efficiency necessary to accommodate the forecasted growth of the system, we need the system to move to one of air traffic management, instead of air traffic control, and to one that is more aircraft centric, taking advantage of cockpit technology and automation.

Historically, continued advancements in technology and automation have indeed reduced the opportunity for human error and increased the efficiency of the crew.

It goes back to the 1940s with pressurized aircraft and stall warnings;  the ‘50s with Instrument Landing Systems, VHF Omnidirectional Receivers, autopilots and radar;  then the advent of the jet engine, and more recently, Traffic Collision Avoidance Systems, Ground Proximity Warning Systems, auto land capability, and Required Navigation Performance/ Area Navigation.  Each new piece gave us advances in safety and, in most cases, efficiency.

Historically, continued advancements in technology and automation have indeed reduced the opportunity for human error and increased the efficiency of the crew.

Today, we’ve had a great run since 2009, the year of the last fatal, U.S. commercial aviation accident.  There were human factor and automation interaction issues present in that accident, as well as AF447 and Asiana Airlines flight 214 at SFO, all of which seemed to have given everyone a bit of a pause, to the extent that the 2013 report from the FAA’s Performance-based Aviation Rulemaking Committee, Flight Deck Automation working group, stated that “pilots sometime rely too much on automated systems”.

That’s a tough statement to swallow.  I could understand it better if the intent was to suggest that pilots rely too much on automated systems that have failed or pilots rely too much on automation systems they don’t understand.  Indeed, there have been several close calls in recent years that continue to confirm the need for using automation.

So, I am encouraged to see companies like Airbus develop project Vahana and Boeing acquire Aurora Flight Sciences.  Not because I think commercial passenger aircraft should be autonomous, but because those developments signal further automation advances for aviation in general, including positive aircraft control and autonomy.

At Rockwell Collins, we continue to move toward more automation, specifically those which are focused on improving pilot performance and reducing workload in the cockpit.  So our R&D includes sophisticated flight path displays in different axes, intuitive touch screens, and enhanced vision, synthetic vision and combined vision technologies.  In our view, these automation features have substantial potential to improve performance and increase safety and efficiency.

We also are continuing to focus on a “head up, eyes out” automation philosophy because we think head up guidance systems provide more safety and efficiency benefits as compared to heads down displays.

We’re investing in voice recognition technologies on the flight deck, as well as autonomy and operations with a reduced crew.

And, finally, we see enhanced safety and efficiency benefits from live, virtual constructive, or LVC training.

Looking ahead, we’ll also see the air traffic control system itself move toward more automation. The FAA’s rollout of Datacomm, a digital messaging capability that reduces verbal communication errors,  has been very successful.  To date, according to the FAA’s website, 55 airports have Datacomm services and the airlines have asked for seven more airports to become part of the program.

Likewise, the introduction of Automatic Dependent Surveillance – Broadcast should facilitate better efficiency from spacing reductions and more direct routing.  With ADS-B, I think we will quickly see innovations in flight path management that will reduce arrival and in-trail separation as well as strategies for maintaining Visual Meteorological Condition arrival rates in marginal Instrument Meteorological Conditions.  These types of procedures may require pilots to relinquish control to the automation.

Looking ahead, we’ll also see the air traffic control system itself move toward more automation.

Looking farther ahead, there are four areas of aviation that continue to need our focus: certification, particularly around fail-safe modes; Human to Machine Interface (or HMI);  training;  and data and collaboration.

Certification has become increasingly complex, almost unmanageable for new aircraft and new automation, so we need to embrace concepts like formal methods to help us better address certification issues.

It has been said that part of the reason certification has become so complex is we have an aviation system with increasing levels of automation, but not fully automated.  It is also still largely human-centric, whether you’re looking at it from the pilot or air traffic controller point of view.  This makes the introduction of new automation more complex than it probably otherwise would be if the system was fully automated.

Regardless, we also need to continue to emphasize redundancy, integrity and failure modes in the certification process…..particularly fail-safe modes and the ability of the pilot to quickly recognize and react appropriately to failure modes.

Training will always be a balance between too much and not enough.  The latest accidents have highlighted several training areas that needed reinforcing.

It is the areas of HMI and data and collaboration where we will need to place our primary focus if we are to realize further safety and efficiency gains from increasing automation.

At Rockwell Collins, we are constantly running pilots through our advanced flight deck lab seeking their feedback on new technologies, displays, information and its presentation to the crew, and even knob placement or touchscreen feel.  This will become increasingly important for automation involving positive aircraft control.

With respect to data and collaboration, this new generation of aircraft and their systems are capable of putting out far more data than we’ve ever had before.  A single Boeing 787 Dreamliner on a single roundtrip international flight puts out a terabyte of data.  We will need to learn how to harvest the data and collaborate, or work together, to ensure we are operating in a manner most conducive to safety and efficiency.

More than any other agency, the FAA collaborates best with industry. The best example of this is the Commercial Aviation Safety Team, CAST, which was formed in the early 1990s and is comprised of industry-government professionals.  By working together – partnering – industry and the FAA reduced the commercial fatal accident rate by more than 80% in less than 10 years.  And CAST continues to do great work today by adopting a proactive, data-driven approach using tools such as the Aviation Safety Information Analysis and Sharing (ASIAS) program to detect emerging risks.

In this new era of ever-increasing automation and autonomy, we’ll need even greater collaboration and partnering between government regulators and industry, in all modes of transportation.   The enforcement arm, at least in transportation, will have to take a back seat to partnership if we are to continue moving forward.   We always want that stick when needed, but a just safety culture demands more collaboration.

More than any other agency, the FAA collaborates best with industry.

I’ll close with some related thoughts about the automobile industry.  Within just the past five years, automation such as rear-view cameras, head-up displays and lane-keeping assist have made considerable inroads in our sister mode of transportation.

The auto industry appears to have fully embraced a path towards more automated, even driverless or autonomous vehicles.

With greater than 30,000 deaths per year and increased traffic congestion, the potential safety and efficiency benefits are simply too big to ignore.  In some respects, highway safety is where aviation safety was back in the 70s and 80s and today’s technologies offer them a chance to leapfrog to a much greater level of safety.

The highway industry’s exuberance is understandable given their safety record.  Likewise, aviation’s “slow but steady” attitude is understandable given our current safety record.

In both cases, autos and airplanes, we need to keep the lessons we have learned from aviation at top of mind:  certification, HMI, training, and data and collaboration.

So I’ll close where I started… with the same message I gave to a crowd of aviation litigators a decade ago.  I know you have your clients and I know the ethical obligations of the profession.

But you are also part of the aviation safety culture.

I’m an aviator.  And I’m a big proponent of the pilot and air traffic control community and the benefits they have contributed to aviation over many decades.  But, aviation and ATC need to continue their march toward increasing automation, and developing autonomous capability.  The technology is broadly more accurate and less prone to errors.

And that is why I think we need to continue down this path of automation.

And YOU need to help keep it moving in that direction.

Thanks again for inviting me to speak with you today.

Fly safe.

Related products

Posted by Rockwell Collins