The guiding principle of air traffic control is that safety is paramount. NATS requires reliable estimates of the extremely low risk of collision between aircraft in its airspace.
A specific problem studied by the NATS Operational Analysis department concerns risk estimation in procedural airspace – in which for practical reasons there is little ground-based surveillance – where aircraft separation is based on precise position reports from the aircraft.
Since no navigation system is perfect it is necessary to base a risk estimate on models of the deviations of aircraft from planned positions. Deviations are usually very small and pose little or no risk, but the very rare larger deviations potentially pose a greater risk. Reliably modelling such rare events is fundamentally difficult, with the risk estimate being very sensitive to the way data is fitted and extrapolated using, for example, maximum likelihood techniques.
The Smith Institute helped NATS to develop a new mathematical approach to calculating air traffic safety.
NATS commissioned the Smith Institute to undertake a case study and carry out a review of alternative approaches to risk estimation.
The Smith Institute critically evaluated and provided a deeper understanding of the analysis methods previously being used by NATS for calculating the collision risk of procedurally separated aircraft.