Persistent Drone Light Shows: Petri Net Models, Performance Evaluation and Resource Requirements

While drone light shows delight audiences, they are inherently transient, providing too brief moments of illuminated joy. Such shows are not alone in their limited duration, as constraints on flight time are a primary feature of UAV systems. Cognizant of the limitations of the show coordination software (SCS), we propose and study a concept enabling increased duration of operation for drone light show systems and strive toward persistence. The central idea is to hand-off the sky from a weary fleet of UAVs to a fully charged fleet and repeat ad infinitum. Each such fleet is supported by a collection of resources which we refer to as a leaf, including extra batteries and chargers so that the fleet can promptly replenish and be available to return to the sky. We develop a decision-free Petri net model for a two-leaf system and, given an initial number of resources in each leaf, conduct performance evaluation. Given a target coverage for the sky as one fleet available per unit time, we obtain the minimum number of resources required in each leaf. These results are generalized to n-leaf systems. The results are applied to assess the costs of deploying a persistent show for three categories of drones (discount-tier, mid-tier, and premium-tier). Our results enable us to provide a side-by-side comparison of the resources and costs of persistence for drones from these categories.