TY - GEN
T1 - Safe flocking in spite of actuator faults
AU - Johnson, Taylor
AU - Mitra, Sayan
PY - 2010
Y1 - 2010
N2 - The safe flocking problem requires a collection of N mobile agents to (a) converge to and maintain an equi-spaced lattice formation, (b) arrive at a destination, and (c) always maintain a minimum safe separation. Safe flocking in Euclidean spaces is a well-studied and difficult coordination problem. Motivated by real-world deployment of multi-agent systems, this paper studies one-dimensional safe flocking, where agents are afflicted by actuator faults. An actuator fault is a new type of failure that causes an affected agent to be stuck moving with an arbitrary velocity. In this setting, first, a self-stabilizing solution for the problem is presented. This relies on a failure detector for actuator faults. Next, it is shown that certain actuator faults cannot be detected, while others may require O(N) time for detection. Finally, a simple failure detector that achieves the latter bound is presented. Several simulation results are presented for illustrating the effects of failures on the progress towards flocking.
AB - The safe flocking problem requires a collection of N mobile agents to (a) converge to and maintain an equi-spaced lattice formation, (b) arrive at a destination, and (c) always maintain a minimum safe separation. Safe flocking in Euclidean spaces is a well-studied and difficult coordination problem. Motivated by real-world deployment of multi-agent systems, this paper studies one-dimensional safe flocking, where agents are afflicted by actuator faults. An actuator fault is a new type of failure that causes an affected agent to be stuck moving with an arbitrary velocity. In this setting, first, a self-stabilizing solution for the problem is presented. This relies on a failure detector for actuator faults. Next, it is shown that certain actuator faults cannot be detected, while others may require O(N) time for detection. Finally, a simple failure detector that achieves the latter bound is presented. Several simulation results are presented for illustrating the effects of failures on the progress towards flocking.
KW - failure detector
KW - flocking
KW - safety
KW - stabilization
KW - swarming
UR - http://www.scopus.com/inward/record.url?scp=78249239499&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=78249239499&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-16023-3_45
DO - 10.1007/978-3-642-16023-3_45
M3 - Conference contribution
AN - SCOPUS:78249239499
SN - 3642160220
SN - 9783642160226
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 588
EP - 602
BT - Stabilization, Safety, and Security of Distributed Systems - 12th International Symposium, SSS 2010, Proceedings
T2 - 12th International Symposium on Stabilization, Safety, and Security of Distributed Systems, SSS 2010
Y2 - 20 September 2010 through 22 September 2010
ER -