TY - GEN
T1 - Self-organized Chain Formation of Nano-Drones in an Open Space
AU - Barciś, Agata
AU - Barciś, Michał
AU - Natalizio, Enrico
AU - Ferrante, Eliseo
N1 - Publisher Copyright:
© 2022, Springer Nature Switzerland AG.
PY - 2022
Y1 - 2022
N2 - We propose a method for the chain formation of multiple agents in an open space. Chaining can be considered as a building block for several application scenarios, including exploration, maintaining connectivity, or path formation. The proposed method was designed for a very sensing and computationally constrained robot platform, more specifically for nano-drones as they offer advantages in applications in tight spaces or in the proximity of people. To enable portability to a real platform, the method relies on a range and bearing sensing model with a limited field of view that is susceptible to occlusions, which was implemented both in simulation as well as on the real robot through a camera coupled with LEDs. We analyze the method in the simulation-based study. We show that the method works even in presence of noise in sensing and actuation, which rather than being harmful to the chaining performance has a positive effect. We analyze the performance in terms of quality of final chain formation, and speed of convergence, and how these two are affected by increasing swarm size. Finally, we present its practical feasibility in a robotic proof-of-concept featuring nano-drones.
AB - We propose a method for the chain formation of multiple agents in an open space. Chaining can be considered as a building block for several application scenarios, including exploration, maintaining connectivity, or path formation. The proposed method was designed for a very sensing and computationally constrained robot platform, more specifically for nano-drones as they offer advantages in applications in tight spaces or in the proximity of people. To enable portability to a real platform, the method relies on a range and bearing sensing model with a limited field of view that is susceptible to occlusions, which was implemented both in simulation as well as on the real robot through a camera coupled with LEDs. We analyze the method in the simulation-based study. We show that the method works even in presence of noise in sensing and actuation, which rather than being harmful to the chaining performance has a positive effect. We analyze the performance in terms of quality of final chain formation, and speed of convergence, and how these two are affected by increasing swarm size. Finally, we present its practical feasibility in a robotic proof-of-concept featuring nano-drones.
UR - http://www.scopus.com/inward/record.url?scp=85142753035&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85142753035&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-20176-9_18
DO - 10.1007/978-3-031-20176-9_18
M3 - Conference contribution
AN - SCOPUS:85142753035
SN - 9783031201752
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 222
EP - 233
BT - Swarm Intelligence - 13th International Conference, ANTS 2022, Proceedings
A2 - Dorigo, Marco
A2 - Strobel, Volker
A2 - Camacho-Villalón, Christian
A2 - Hamann, Heiko
A2 - Hamann, Heiko
A2 - López-Ibáñez, Manuel
A2 - García-Nieto, José
A2 - Engelbrecht, Andries
A2 - Pinciroli, Carlo
PB - Springer Science and Business Media Deutschland GmbH
T2 - 13th International Conference on Swarm Intelligence, ANTS 2022
Y2 - 2 November 2022 through 4 November 2022
ER -