Pilot Training: complexity of the human-technology sinergy. The Humai Research perspective.
Luca Lisci: Welcome Giacomo and Fabrizio. I am delighted to have you here at this edition of IDA. We met at the DSA Summit in Milan in June 2023, and since then, I have had the pleasure of getting to know you better and joining the foundation of the HUMAI Research project, of which you are the first promoters. I invite you to introduce the project to our readers on this occasion. However, we would be grateful if you could briefly introduce yourselves.
Giacomo Belloni: I am a pilot and airline Captain. I have been involved in advanced aeronautical training for many years. I have also worked in academia and university as a researcher on development and research projects in this field.
Fabrizio Interlandi: I have been an airline pilot and Captain for many years, and in addition to working on advanced training. I collaborate with some universities on research projects, and with the HUMAI Research team, Giacomo and I work as a trait union between the academic and aeronautical worlds.
Luca Lisci: Let’s start from the context: what’s the situation of commercial aviation today?
Giacomo Belloni: Today, the commercial aviation environment is characterised by an elevated grade of complexity that often shows unprecedented and unpredictable dynamics for which we are never prepared enough. The ways, the modalities and the effects are different every time and vary from those we are used to. Furthermore, it is characterised by impressive dynamism. It transforms and alters reality without ceasing.
As Taleb would say – as unfortunately has often and amply been demonstrated in commercial aviation as well – when complexity comes up negatively, the effects are those of the Black Swan: unpredictability and devastating impact. Furthermore, and this is its third characteristic, after the event has occurred, we find – but only in hindsight – simple explanations that the event was nevertheless predictable and less random. This means the tendency is to trivialise its polymorphism and remain unprepared to manage the next event.
Luca Lisci: So, how do we approach and manage complexity in your vision?
Giacomo Belloni: It is essential to constantly evaluate established methodologies to understand if we are leaving something behind and what is transforming in front of us. All the complex systems, including aeronautical ones, to contain and reduce complexity to manageable variables tend to be close to external inputs. While this helps manage complexities, on the other hand, it can limit circular comparison with different realities, impacting development and innovation.
Fabrizio Interlandi: The aeronautical training system, precisely because it is complex, is exposed to this threat, and it is essential to adopt dynamic approaches to remain in line with constantly evolving needs and challenges.
The introduction of evidence-based training about ten years ago is a clear example of how the system has reacted to the risk of remaining rooted in notions, theories, and methodologies that were not aligned with the needs that have evolved over time.
To ensure that the system is constantly keeping up with the speed at which the world is evolving, it is essential to have points of contact with scholars and the research world. The interaction, for example, with teaching and learning experts and SMEs of the academic world, contributes to the identification and development of training techniques that take into account emerging technologies, generational changes and current problems. This continuous and constructive confrontation allows aeronautical training to identify the most suitable methodologies to prepare pilots to face constantly evolving challenges with adequate competence and resilience. Change must be addressed systematically and structured, seeking collaboration and comparison with other realities.
Giacomo Belloni: The commercial aviation system represents one of the highest levels of complexity in the contemporary world. Moreover, it is a system that is becoming more articulated and varied daily. The impressive efficiency required today by commercial aviation systems leaves no room for errors or inaccuracies. It is an extraordinarily multifaceted system that must allow as many aircraft as possible to be in the sky at the same time. Imagine the level of precision for adequate horizontal and vertical separation or margins for climbing or descending with no conflicts that must be guaranteed to the tens of thousands of aircraft flying in today’s congested skies.
Fabrizio Interlandi: The high level of technology in cockpits today allows us to maintain an impressive precision of trajectories so that there can be more and more aircraft in the skies to meet the increasing demand.
Luca Lisci: How are pilots trained to deal with this high level of system complexity?
Fabrizio Interlandi: Evidence-based training is the most innovative and advanced training methodology today, used by many of the most advanced airlines to train their crews. Its objective is identifying, developing, and evaluating the competencies required for pilots to operate safely, effectively, and efficiently.
Giacomo Belloni: For decades, pilot training has been based on the constant repetition of critical manoeuvres and the management of a limited number of predictable and well-codified failures, even though it was clear for a long that accidents occur for different, unprecedented causes, for which traditional training no longer offered practical solutions.
As IATA wrote in 2013, analysis of the past few decades has clearly shown that the human factor and related errors have been the determining cause of between 70% and 80% of accidents in the aviation sector.
Over time, the inadequacy of traditional methodologies becomes more and more amplified. The only way to deal with the present day is to train pilots to be resilient and capable of managing the unpredictable, the Black Swan.
Moreover, today’s high degree of cockpit technologisation and automation presents a further complication.
Luca Lisci: Yet, I can’t imagine how this complexity can be managed without advanced technologies.
Giacomo Belloni: Technology in the cockpit has always been and will continue to be fundamental both for the industry and for the pilots who use it. The expected increase in complexity and the need to make safe, efficient, and economical decisions in a short time, capable of taking into account more and more elements and variables, make the introduction of increasingly cutting-edge technologies, such as artificial intelligence, into the cockpit inevitable today.
The more complex, articulated and multifaceted systems become, the greater the demand for innovative technologies to deploy, monitor and manage them safely and efficiently and to enable pilots to manage the increasing level of complications. It inevitably brings advantages that lead to unwanted consequences if not well handled by updated and practical training. Technology must be understood and well metabolised by the utilisers. As has happened in the past, technology has often given rise to new problems and concerns. Errors in highly automated aircraft are increasingly linked to the adverse effects of automation on the pilot’s situation awareness, workload management, and ability to revert to manual control. Adaptive systems should be designed according to a human-centred automation philosophy to be effective. In a highly sophisticated environment, the pilot has often proven to be the weak link because, if not adequately trained with methodologies suited to complexity, the human-machine relationship increases its exposure to error.
I’m not talking only about specific training for the use of advanced systems but of methodologies and philosophies that consider the human factor in the technical and non-technical management of a highly technological machine in a hyper-complex environment. Furthermore, technology must be developed with the constructive involvement of all stakeholders, starting from the pilots.
Luca Lisci: And this point is the foundation of the HUMAI Research project.
Fabrizio Interlandi: Yes. We want to study and understand how to train pilots in its use so that artificial intelligence becomes a tool and a valid support to enhance their competencies.
As we have seen in the past, the introduction of new technologies, while on the one hand, increases the efficiency of the system, on the other hand also increases its complexity; in moments of high workload for its operators, the management of a high degree of complexity can sometimes become a problem instead of an aid. Specifically, together with the researchers of HUMAI Research, we want to understand how advanced technologies, including artificial intelligence, can be effectively introduced into aeronautical training for a more effective preparation of pilots to face the multitude of contemporary scenarios.
Giacomo Belloni: Evidence-based training has filled a decades-long inadequacy of adaptation to modernity. The challenge is to continue to develop advanced technological implements and training methodologies to face the near future. Like evidence-based training, artificial intelligence is based on an approach based on objective data. This common element creates a close connection between these two realities. Therefore, it is reasonable to expect a natural integration between evidence-based training and AI in the near future. Artificial intelligence is a transformative force in aeronautical training, as it can offer personalised, efficient, and data-driven approaches to improve pilot skills and competencies.
EASA, the European regulator, has published documents underlining the importance of artificial intelligence in aeronautical training. The emphasis of many high-level international aeronautical bodies and entities on the research and development of artificial intelligence reflects the high commitment and interest to harnessing the potential of artificial intelligence to improve aeronautical training and, by extension, aviation safety.
The future of aeronautical training is undoubtedly intertwined with the growth and implementation of AI technologies.
As HUMAI researchers, we aim to understand how to make it usable, friendly and entirely accepted by its operators, investigating how and in which way to meet their needs. It must be developed ethically and together with the final utilisers.
Luca Lisci: I would say that this is a project that is fully in the spirit of the times: advanced technologies, complexity management, and the centrality of the human being, all applied to one of the most relevant global issues of our time: aviation.
Giacomo, Fabrizio, thank you for bringing your contribution to this edition of IDA. Will we see you again at the DSA Summit in Valencia next June?
Fabrizio Interlandi: We can’t wait!
Giacomo Belloni: It will be an exciting opportunity for us to extend the HUMAI project to those interested in developing solutions for Training in Complexity. Decision Science Alliance is the ideal setting in which to spread our initiative.