ICRAT 2020 Morphing STARs vs drones and weather in TMA Henrik Hardell Communications and Transport Systems, ITN, Link¨ oping University, Norrk¨ oping, Sweden Procedure Design Unit, Luftfartsverket (LFV), Norrk¨ oping, Sweden Anastasia Lemetti, Tatiana Polishchuk and Valentin Polishchuk Communications and Transport Systems, ITN, Link¨ oping University, Norrk¨ oping, Sweden Vishwanath Bulusu Crown Consulting Inc., Moffett Field, California, USA Cal Unmanned Lab, University of California, Berkeley, California, USA Enric Royo The School of Industrial, Aerospace and Audiovisual Engineering of Terrassa (ESEIAAT), TU Barcelona, Spain Abstract—We present an optimization framework for comput- ing STARs that slowly change over time, while always avoiding a set of moving obstacles in TMA. The framework is applied to two types of obstacles: a drone intruder and hazardous weather. We demonstrate the output of our algorithms on synthesized drone intrusion incidents and real storm cells in Stockholm Arlanda terminal area. Keywords—TMA route design; ATM and UTM; Weather avoidance I. I NTRODUCTION Drones interfering with aircraft on approach are reported with alarming frequency. Such interference poses serious risk to development of the emerging drone industry, as it negatively impacts public acceptance of UAV (Unmanned Aerial Vehicle) operations. Economic damage and passenger discomfort is es- pecially aggravated when drone sighting leads to cancellation of flights through the whole terminal airspace and prolonged freeze of airport operations. We are thus motivated to consider closing only that part of the airspace where the drone is physically present and recompute the arrival tree so that the aircraft safely land, avoiding the drone-affected part. The challenge is that the drone is moving, implying that the drone-impacted area is a moving obstacle. In addition, since the drone may be spotted only once, its uncertainty region grows, meaning that the drone obstacle not only moves but also may expand with time. This makes the drone deconfliction similar to weather avoidance i.e. escaping both kinds of obstacles fits into the same abstract model. In this paper we extend the optimization framework from [1] for constructing operationally feasible obstacle-avoiding static standard arrival routes (STAR), to handle obstacles changing This research is supported by the SESAR Joint Undertaking under the Euro- pean Union’s Horizon 2020 research and innovation programme under grant agreement No 783287. It is also partially supported by Transportstyrelsen, Trafikverket and in-kind participation of LFV. with time and to output morphing STARs that never change abruptly (so controllers and pilots may easily see how the STARs evolve over time). Our mathematical model for the morphing STARs works with any type of abstract obstacles. To illustrate it, we present experimental results of application of the morphing STARs to drone deconfliction and to weather avoidance in the terminal maneuvering area (TMA) of Stock- holm Arlanda airport. While the drone intrusion scenario is synthesized, for the weather data we took a real storm. The paper is organized as follows. The remainder of this section reviews related work. Section II formalizes the model and Section III casts our problem as an integer linear program (IP). In Section IV we apply the IP to drone and weather avoidance in Stockholm Arlanda TMA. Section V concludes the paper. A. Related work Technologies proposed to detect, track and shoot down drones include radars, cameras, eagles [2], counter-drone UAVs [3] and jamming [4] – a comprehensive survey can be found, e.g., in the thesis [5]. Still, a commonly accepted and implementable solution has yet to be developed (e.g., the Blue Ribbon Task Force report [6] on UAS mitigation at airports unfortunately offers ”disconcerting” [7] conclusions). The governments are not ready to deploy the offered solutions primarily because technologies for multi-aviation airspace must first go through the cycle of validation and certification. Furthermore, the actual handling of the drone intruder is up to the police, who would appreciate information sharing both with ATM and UTM (or the future combined system). It is also necessary to correlate the detected drone with what UTM knows, in order to distinguish between a lost drone of a good citizen and a malicious intent to fly unnoticed. Impact of deep convection and thunderstorms is also subject to ongoing research, e.g. Steiner et al. [8], [9] and Song et al. [10] investigated its implication both on the en-route