Aerospace America publishes a wide variety of opinions relevant to the future of aerospace. The views expressed are those of the author(s) and do not necessarily reflect those of our publisher, AIAA.
The aviation industry has long been chasing the advantages of automation, which reduces a flight crew’s workload to improve safety. More than 100 years ago, Sperry demonstrated the first autopilot, a simple wing leveler. By the 1960s, autopilot not only handled all three axes, it was also integrated with a flight management computer and its output displayed on a flight director. Airliners soon incorporated autothrottles to regulate and balance the flight forces.
The U.S. Air Force began incorporating Auto-GCAS (Ground Collision Avoidance System) into its frontline fighters earlier this century. Pilots initially resented the control the technology removed from them when Auto-GCAS determined controlled flight into terrain was imminent; however, it is now widely accepted, saving over a dozen lives to date and earning the 2018 Collier Trophy, the most esteemed acknowledgement of aerospace advancement. Further illustrating this success, Air Force is now developing a companion: Auto-ACAS (Air Collision Avoidance System).
Early commercial airliners required a crew of five on the flight deck, but as technology and automation matured, the radio operator, the navigator and flight engineer positions were phased out. Today, the potential of automation technology means we are approaching the point when the co-pilot and perhaps even the pilot are no longer necessary and can be replaced with automation.
Several years ago, I had the opportunity to fly an Airbus A350 simulator. This widebody is one of the newest commercial aircraft and features the most advanced automation and computer systems available when it was designed 20 years ago. I am a licensed airline transport pilot and have flown many different types of aircraft, but on that day, I was struck by how the levels of automation reduced my workload. (Plus, I was flying as a single pilot with an instructor in the right seat to flip switches that were beyond my reach.)
I performed one of the most challenging emergency events: an engine failure and fire at the point of rotation. Upon the engine failure, I rotated an additional five degrees of pitch, per the procedure, and used rudder to hold heading. At this point, the instructor told me to let go and just release all the pressure I had put in to hold a safe and stable attitude. In another aircraft, that would result in an immediate loss of control, but in the A350, the computer continued to fly the aircraft as I had left it.
I then called for the memory items and emergency checklist for an engine failure on takeoff; yet, the instructor pointed me to one of the multifunctional displays, the Electronic Centralized Aircraft Monitor. On it was the checklist, and nearly all the items had been marked completed: Thrust on remaining engine to maximum. Affected engine thrust to cutoff. Fuel system, electrical systems, hydraulic systems — all reconfigured. It only prompted me to check the oil pressure on the affected engine and, if it read zero, shut off the engine. Other than that single operation, I could continue to fly the aircraft with no additional required actions other than the obvious reduced performance associated with single-engine operations. Automation greatly reduced and simplified my decisions and increased flight safety.
During a panel at AIAA’s AVIATION Forum in July, Brian Yutko, vice president of product development at Boeing Commercial Airplanes, stated that future air vehicles will include “automation for safety enhancements on crewed airplanes, or autonomy for flying uncrewed airplanes and all kinds of interesting solutions.” Maybe the issue isn’t the abilities of automation or autonomy to safely control the aircraft, but instead public acceptance and political approval. Outside the U.S., there are airlines pursuing single-pilot operations (SPO) on long-range flights that today require multiple crews due to crew rest requirements. The first step toward SPO would be eMCO — extended minimum crew operations — which would require two pilots in the cockpit during departure and arrival, but only one during cruise, with another on ready reserve. Under this arrangement, a rotating two-person crew could handle a long-range flight, rather than the four-person crews that are mandatory today.
Yutko is also chairman of the board and former CEO of Wisk, a wholly-owned Boeing subsidiary that is developing what could become the first fully autonomous, passenger-carrying electric vertical takeoff and landing air taxi in the U.S. Fully autonomous AAM vehicles are a necessity for economic, staffing, and practical reasons. Economically, an operator must have a paying passenger in a seat that otherwise would require a paid pilot. The projected growth of AAM operations could not be met with the projected training supply of qualified pilots, which is already struggling to match airline demand. And practically, an automated system can respond more quickly and safely to the rapidly changing urban flight environment.
As a member on the same AVIATION panel as Yutko, I predicted we’d have large, completely autonomous cargo aircraft operating by 2050 — but not in the Western world. Rather, I see widebody cargo aircraft without onboard pilots flying wherever there is a large need for moving cargo, as well as a political and policy environment that will permit autonomous flight. I believe that with decades of experience and public acceptance of fully autonomous on-demand taxies on the ground, plus autonomous AAM, there will be eventual acceptance of larger autonomous air vehicles.
However, the Air Line Pilots Association has consistently opposed the introduction of SPO, in commercial aviation, citing safety concerns. The association says the presence of two pilots on the flight deck provides a crucial layer of safety and redundancy as it offers two trained and rested pilots on the flight deck. Part of the concern is that current and projected automated system technologies do not offer an acceptable or equivalent level of safety compared to the traditional two-pilot model.
For an airliner to be acceptable for SPO, it must be able to deal with the remote possibility of incapacitation of the single pilot. In such an occurrence, it then must be able to safely default to a fully autonomous mode and meet the same safety levels we expect. Those current thresholds are failure rates of one in a billion for air transport and one in a million for single-engine aircraft.
I’ve also had occasion to fly a demonstration of the Garmin Home Safe, also known as Garmin Autoland, which won the Collier Trophy in 2020. Upon pressing a button the system took over all control of the aircraft, selected an appropriate airport, made necessary announcements via radio, lowered the flaps and gear, made a reasonable landing on the centerline of the runway, brought the aircraft to a stop, shut down the engine, and presented instructions to passengers on how to open the door and egress the aircraft.
Today, I don’t believe we yet have public acceptance of any type of autonomous vehicle, ground or air, as those autonomous ground taxis are still very much a novelty. However, with more automation being incorporated into nearly everything and more autonomous vehicles being developed and deployed worldwide, there will be a growing acceptance of driverless cars and trucks and eventually pilotless air vehicles. The progress is inevitable as we are on a path to get to fully automated flying vehicles in the next several decades.
