Unravelling complexity in air traffic management
Andrew Cook, University of Westminster, has a passion for unravelling complex systems and for finding ways to improve them. It is this curiosity that initially drew him to air traffic management and SESAR where he has been active in a number of exploratory research projects. In 2017, he became a member of the SESAR 2020 Scientific Committee, providing recommendations on the programme’s activities. In this article, he describes the road that took him from enzymes to aircraft, and how complexity science can be used to improve ATM operations and performance.
What drew you to research and more specifically to ATM research?
After studying natural sciences as an undergraduate, my postgraduate research was in the physiological sciences. More specifically, I was working on the design of new enzyme inhibitors as drugs to control diseases such as HIV and hypertension. At the same time, I supplemented my student grant by working on various transport-related research projects, usually carrying out surveys at unearthly times of the day at ferry ports and airports.
These activities were all underpinned by a curiosity for understanding how systems work, and interventions that can improve them. The systems progressively became bigger: from atoms, to enzymes, to aircraft. The lure of ATM research was partly that man-made systems appeared somewhat easier to improve through design, as compared with less well understood biochemical systems. However, I now realise that complexity (and emergence) is our friend, and foe, at both extremes of scale.
What was the last “Eureka” moment you experienced?
Rather than a given moment, I’d like to identify a type of moment, one that never ceases to motivate my passion for the next piece of research we’re working on. This is when a model gives an unexpected result. It is clearly ‘wrong’, and you duly set out to fix the model, but can’t find the problem. Then, sooner or later, the realisation pops into your head, that it was your understanding that was wrong, and not the model (OK, it often is actually the model, but those experiences are rarely the exciting ones!). It is so rewarding to reach a counterintuitive result that gives greater insights into the problem than simply confirming original hypotheses.
If one can have an extended ‘Eureka!’ experience, a slower burn for me has been an increasing engagement in complexity science, from understanding the rather broad church of its principles, to being engaged in the development of new models that teach us about the ATM system. Clearly, we still have a long way to go, but these techniques are offering distinct new insights.
What are the challenges facing air traffic management?
I would like to identify three interconnected challenges. The first is intermodality. It is clear that as the focus moves increasingly towards door-to-door solutions and integrated mobility management, transport sectors will be less able to operate independently. In many respects, air transport in general, and ATM in particular, is significantly ahead of other transport modes. In some others, there is still work to be done (even without drones being added to the picture). Sometimes, air transport does not feel as efficient and resilient as we, the passenger, would like it to be, and extending the scope of mobility integration is clearly going to be even more of a challenge.
Second, are the further pressures arising from greater focus on performance-driven change. Even with the relatively few (current) key performance indicators (KPIs) in the Single European Sky (SES) Performance Scheme, many stakeholders report difficulties mapping the trade-offs between them. Building improved tools to quantify these trade-offs and make better, evidence-based decisions will become more difficult still as we extend the scope to include passenger mobility at the door-to-door scale. Third, and coupled to the first two challenges, is how we extend our operational and performance models to better embrace increasingly complex, more highly coupled, socio-technical systems of systems.
How do you see SESAR meeting those challenges?
To consider the latter two first. There has been a lot of excellent work in performance assessment to date, both in the industrial/applied and exploratory research arenas of SESAR, and much splendid work continues. At the same time, I believe there are good opportunities ahead for increasing coordination across these initiatives, to maximise the benefits delivered to the stakeholders. Some of these opportunities are currently being actively explored by the SJU, with support from the Scientific Committee. Also, exploratory research projects are being launched further investigating various aspects of system coupling and stakeholder impacts, for which foundational concepts and early work was established in SESAR 1, including the ComplexWorld network. Although these types of investigation remain largely as fundamental research, it is exciting to see some applied ideas emerging from them already, and the continued recognition of their importance to ATM.
As regards tackling the intermodal context, this exciting challenge lies very much ahead of us. For example, integrating digital consumers and autonomous vehicles into door-to-door mobility. ‘Paradigm shift’ is perhaps an overused expression, but achieving ACARE’s goals by 2050 means that we will certainly have to think in rather different ways about service delivery and KPI trade-offs, with increased pressure on data accessibility and security.
How do you feel your expertise and participation in the SESAR Scientific Committee will contribute to supporting the SESAR vision?
It’s rewarding being a member of such a diverse team, with a clear commitment to supporting the work of SESAR and being given a free hand to input into the research agenda. We’ve already established a number of task forces with explicit objectives helping to shape the research vision. The Committee has been selected to cover a range of skills and experiences across the ten members. I hope that my experience of nearly 30 years (which makes me feel rather old!) in a range of transport sectors, not only wider air transport, often with a passenger-centric focus, will contribute to the thinking of the steps we must take towards the long-term integration and optimisation of European mobility, increasingly extending our focus beyond the gate-to-gate context. I have had the privilege of working on a number of performance assessment projects, deploying both classical approaches and complexity science techniques to developing improved solutions and tools for stakeholders, including passengers and the environment. I hope that this experience may contribute something to the work of the Committee, and how (air) transport collectively aligns long-terms goals with research and new regulatory requirements.
Can you mention a couple of projects in the current exploratory research portfolio that you feel really stand out and why?
BigData4ATM explores how to analyse and integrate a variety of emerging passenger-centric data to fill information gaps in sources traditionally used for socio-economic and behavioural studies. Its goal is to enable better informed decisions, taking into account their full impact on the passenger. It has successfully demonstrated how (expensive/inaccessible) traditional data sources on passenger behaviour may be replaced/supplemented by big data, e.g. from geo-located tweets and public transport smartcards.
INTUIT explores the potential of visual analytics and machine learning techniques to develop new decision-support tools for ATM performance monitoring and management. The project is structured around several case studies, each addressing a problem where the trade-offs between ATM KPAs and the cause-effect relationships between performance drivers and indicators are not sufficiently understood. The first case study focused on understanding the trade-off between fuel costs, cost of delay and congestion costs in airline route choices.
Both projects are thus helping to pave the way towards solutions to the challenges I mentioned earlier, such as door-to-door integration and performance trade-offs. I can recommend the fine papers available on the SESAR Innovation Days website.
Read about the SESAR JU Scientific Committee
Read about Committee member Rita Markovits-Somogyi