Guest Post by Steve Miller (first posted on CERM ® RISK INSIGHTS – reposted here with permission)
Commercial aviation is undergoing constant technological change which, although apparent to the aviation community, is little noticed or understood by the flying public. These changes are driving the industry inexorably toward increasing automation for reasons of both safety and cost reduction, the ultimate long term goal being to produce the “totally autonomous” aircraft.
This long term process entails significant risk in many areas, some of the principal ones being identified in the following sections. Aside from the few suggestions mentioned here, this industry will need to develop formal mitigation strategies for all these areas of concern, as well as others that may eventually arise.
- RATE OF TECHNOLOGICAL CHANGES: This process involves the integration of automation into the normally “human piloted” aircraft. In the long run, this means replacing pilot-centric instrumentation and controls with automated systems which accomplish the same thing. Over time, this will require increasing levels of system capability using Artificial Intelligence (AI). Today’s aircraft are already relatively “smart” and have made significant strides in this direction. Each successive step in the implementation of technological improvement must be built upon the foundation of the previous one to ensure that safety is not inadvertently compromised.
RISK: Lack of acceptance by the flying public, possible corporate losses.
- TIMING OF THE TRANSITION TO AUTONOMOUS FLIGHT: Today we have manned aircraft control with some level of automation. We also have Unmanned Aerial Vehicles (UAV’s). We are now beginning to effectively integrate the two, the ultimate goal being to reach the point of fully autonomous flight. Both the industry and the flying public need time to be absolutely certain that it works reliably to come to fruition. Much simulation and flight testing will be required, covering every conceivable “what-if” to avoid disaster! This is a multi-decade process, to say the least!
RISK: Reaching this point too soon could compromise safety, potential for huge corporate losses.
- AIRCRAFT DESIGN: The gradual implementation of increasing automation presents a difficult design challenge. The process begins with enhancements to aircraft control while retaining flight crew authority. As time progresses, more authority will shift to the aircraft itself. Ultimately all cockpit display/control functions will vanish, as will the last vestiges of any monitoring by human pilots or technicians. These changing requirements will need to be based on the proven success of the previous phase. Both the aviation industry and the flying public will need to be convinced that it’s safe to proceed to the next phase of development, as evidenced by accident rates and general reliability.
RISK: Cost overruns, schedule delays, loss of competitive advantage.
- AIRSPACE SYSTEM IMPACT: The airspace-aircraft interface will necessarily become increasingly focused on direct air-ground interaction as opposed to today’s pilot-focused system. Preliminary text-based communications are already in use, replacing the “party-line” voice communication which has existed since the early days of aviation. The FAA’s regulatory and rule-making focus will likewise need to transition to a UAV-based philosophy as well, since the human flight crew will be gradually disappearing. The process of removal of the human element will have to keep pace with the changes made by the aviation industry, requiring close cooperation between government agencies and commercial organizations.
RISK: Funding and schedule delays, failure to match industry technological progress.
- FLIGHT CREW IMPACT: This is likely to become a difficult sticking point. The pilot and flight attendant unions will logically oppose the transition to full automation. To some extent, this will ensure that things don’t progress too rapidly, allowing sufficient time to prove that such aircraft are truly ready for “prime time” and to give this side of the industry enough time to gradually reduce its ranks over several decades, avoiding disruption in those careers. Hopefully, other career choices will emerge to replace those lost to this process. These could include design/maintenance of aircraft, avionics and the airspace system. In addition, the airspace system my implement some level of human monitoring as an added layer of safety.
RISK: Union opposition, lack of alternate career opportunities.
- SAFETY AND COST AS DRIVING FORCES: Fully operational autonomous vehicular systems, whether land, sea or air, will be considerably safer than today’s human-operated versions. In the case of an aircraft, autopilots have long been able to accurately and consistently guide it to the runway threshold. Achieving a safe, fully autonomous aircraft, while still a distant future goal, is a logical extension of today’s capability. Considerable cost reduction will also ensue from the removal of pilot-centric instrumentation and controls, as well as the flight crew.
RISK: Accident potential, profitability issues.
- HOW TO DETERMINE WHICH CHANGES TO MAKE AND WHEN: Proceeding towards increasing automation is a tricky business. Too soon, and it may be full of unknown design flaws; too late, and it may lag behind a sharper competitor’s move. The flying public is already “gun shy” of technological changes; just look at the current 737 MAX experience. One needs to be careful! Removing the human element from the cockpit must be done over a sufficiently long period of time to prove to the public that it’s safe, governmental approval notwithstanding!
RISK: Insufficient corporate vision/expertise to make timely decisions.
- STATUS OF THE PROCESS: Existing commercial aircraft, coupled with the airspace system, are capable of automatically completing virtually all phases of flight, from shortly after takeoff to the runway threshold at the destination. They do, however require programming and monitoring by the flight crew, which can take over at any time. In the field of privately owned single-pilot business aircraft, GARMIN has recently introduced a system which allows the passengers to safely land the aircraft with the push of a button in the event of pilot incapacitation. This system is intelligent enough to check the weather, locate an appropriate airport, communicate with air traffic control and bring the aircraft to a stop after executing the approach and landing! In a related effort, several firms such as AIRBUS, SIEMENS, BOEING, UBER and others are actively developing urban air taxi systems based on UAV designs, with several in flight test. Operations are planned at several locations around the world by 2023 to 2030, such as Dallas, Los Angeles, Singapore, Melbourne to name a few.
The trend toward vehicular autonomy has already begun, commercial aviation being among the primary industries involved. It will take both vision and dedication to risk assessment and management to insure a timely and safe outcome.
Steve is a retired EE with an aviation background. He began flying at age 16, has 6000 hours+ and has been a naval aviator flying AD-5N’s, Convair 240’s/440’s for Mohawk Airlines and a fixed base operator doing both flight instruction and corporate flying, including a stint with the Civil Air Patrol.
He has two separate 4-year degrees: A BS in Business from Miami University, Oxford, Ohio and a BSEE from UMASS Dartmouth. He’s worked for organizations both small and large, commercial and military in the areas of management, marketing and design, as well as experience as an avionics technician. His engineering career includes both analog/digital circuit design, printed circuit board layout and project management. He’s worked in several different industries: autopilots (Edo-Aire Mitchell, circa 1970), sonobuoys (Raytheon) and various digital data systems for Motorola, 3COM and others. His current passion is the development of safety enhancing instrumentation for general aviation aircraft. He’s also written articles many years ago predicting the coming of age of autonomous aircraft.