To that end, the global market size reached $32.5 billion in 2023 and should surpass $78.5 billion by 2032. Advances in drone technology, falling costs, growing consumer demand for efficient delivery, and the improvement of regulatory frameworks have all contributed to this growth trend.
This expected growth can end up attributed to several factors, including advancements in drone technology, reduced costs, increasing consumer demand for faster and more efficient deliveries, and the development of supportive regulatory frameworks, such as those of the Federal Aviation Administration (FAA).
As drones work collaboratively in swarms, however, cybersecurity risks are becoming increasingly prominent.Drone swarms rely on data sharing to perform task scheduling, and this highly interconnected communication mode means an attack on one node could trigger a chain reaction across the entire network, akin to the risks faced by hotel chains or supply chain networks.
To address this challenge, a study by Stefano Chiaradonna, Petar Jevtić and Nicolas Lanchiera from the School of Mathematical and Statistical Sciences at Arizona State University, proposed a dynamic percolation model for the aggregate loss distribution for cyber risk of various-scale drone delivery networks.
Modeling Cyberattack
The researchers imposed contagion processes, based on percolation theory, on the drone delivery’s network to model a cyberattack. They accounted for the randomly determined nature of drone delivery network topologies by modeling a random number of drones per cell tower.
.
In particular, researchers investigated two network topologies for drone delivery networks: The drone swarm as a random mesh network and the out-of-range as a random star network.
Furthermore, through probabilistic analysis, researchers characterized the analytical bounds for the mean of loss and the variance of loss via simulation-based methodology to obtain the main aspects of the loss distribution of cyber risk.
Characterizing the financial losses of cyber risk can provide valuable insights for enhancing cybersecurity measures and implementing strategies to mitigate such risks effectively. Additionally, the model employs a simulated Poisson point process to generate the random spatial networks representing the locations of the drones without the need to know exactly the path or physical environment of the drone delivery operations.
By analyzing single-layer and multi-layer models, the study evaluates the upper bounds of losses for drone swarms and turret networks under cyberattacks, considering different system parameters such as signal strength, communication range, and node vulnerability.Simulation results indicate with lower network percolation risks, optimized allocation of defense resources and improved communication protocols can significantly reduce losses; conversely, in cases of high percolation probability, losses tend to increase.
Click here to view the entire paper.

