An Efficient Algorithm for Self-Organized Terminal Arrival in Urban Air Mobility
Josh Bertram, Peng Wei
Abstract
Urban Air Mobility (UAM) is a concept for future air transportation where air taxis move passengers between vertical take-off and landing sites known as vertiports. While some form of a structured airspace is likely, it is expected that UAM systems will be required to deal with high traffic densities and will need to respond to conflicts due to a dynamically changing environment. Due to the difficulty of predicting demand of air taxi usage, it may also be more difficult to predict the required demand on any given vertiport. We explore an airspace design which can regulate the flow of aircraft landing at a vertiport, maintaining the aircraft in sequence until capacity is available at the vertiport. We demonstrate an implementation of the airspace design using a highly-efficient Markov Decision Process (MDP) based algorithm to provide separation and collision avoidance for a UAM terminal arrival sequencing problem. The vertiport maintains basic information about capacity and sequencing, and the aircraft seamlessly perform guidance in a self-organized distributed manner performing conflict avoidance while waiting to land.