Predicting Conflict Free Trajectories Using Supervised Machine Learning, Initial Investigations

This paper presents initial investigations on the prediction of conflict free aircraft trajectories using supervised machine learning. The motivation is to generate trajectory proposals to resolve conflicts based on current practices (imitation learning) as a way to get controller acceptability. The paper explores two distinct approaches. The first one takes a pilot point of view with a flight centred representation of the surrounding traffic, while the second one takes a controller point of view with a sector-based representation of the traffic. In addition, for the first approach, the traffic input is represented by an image going into a convolutional neural network, while in the second it is represented by a list of flights parameters going into a feed forward neural network.The case study addressed is to predict conflict free trajectories with a 5 minutes look-ahead. It relies on recorded traffic data from 2018 from a busy European en-route centre (Maastricht UAC) used to draw a 250k data set. This dataset was split in two 50% sub-sets: one with no change in vertical and/or horizontal dimension, the other with a change (change thresholds of 1000ft and 2NM determined statistically). The performance of both models is compared to a baseline to ensure a learning has been achieved. For the best model (sector based), the median deviations between the prediction and the true future locations are 0.4NM and 23ft "with no change", and 1.3NM and 500ft "with change". These results show that relevant information has been extracted and a mapping between inputs and outputs achieved. However, the prediction error remains quite significant compared to separation standards (5NM, 1000ft).Future work will investigate further both models, in particular analyze error (focus on bad performance cases patterns) and improving them, and later, adding other information (e.g. military areas, meteo).