TY - JOUR
T1 - Demonstrating a Swarm Production lifecycle: A comprehensive multi-robot simulation approach
AU - Avhad, Akshay
AU - Schou, Casper
AU - Arnarson, Halldor
AU - Madsen, Ole
PY - 2025/4
Y1 - 2025/4
N2 - Swarm Production is a structurally self-organising paradigm that aligns with flexible and reconfigurable manufacturing principles to achieve high-variant and changeable volume market demand. The production lifecycle defines an adaptive topology planning phase, a coherent workstation and multi-robot task-allocation & scheduling phase, and a fleet management phase. A Topology Manager system handles the layout optimisation and reconfiguration within the planning phase of Swarm Production. The layout optimisation undergoes recurrence during a production lifecycle and hence, becomes a dynamic layout planning problem. A Swarm Manager system executes production scheduling and multi-robot fleet management tasks based on the optimised layout in the Topology Manager. The exhibition of an entire lifecycle is crucial to demonstrate the capability of this paradigm and study the stochastic nature of production output due to the changing topologies. A software-in-the-loop simulation for Swarm Production demonstrates multiple scenarios executing multiple production orders with different product mixes. This research work also includes integrating all the systems to form a production suite. The work concludes with quantitative data acquired from the scenario-specific simulations and a formal analysis based on the results. The research contributes as a first full factory demonstration and a potential test bed for upcoming research undertakings within Swarm Production.
AB - Swarm Production is a structurally self-organising paradigm that aligns with flexible and reconfigurable manufacturing principles to achieve high-variant and changeable volume market demand. The production lifecycle defines an adaptive topology planning phase, a coherent workstation and multi-robot task-allocation & scheduling phase, and a fleet management phase. A Topology Manager system handles the layout optimisation and reconfiguration within the planning phase of Swarm Production. The layout optimisation undergoes recurrence during a production lifecycle and hence, becomes a dynamic layout planning problem. A Swarm Manager system executes production scheduling and multi-robot fleet management tasks based on the optimised layout in the Topology Manager. The exhibition of an entire lifecycle is crucial to demonstrate the capability of this paradigm and study the stochastic nature of production output due to the changing topologies. A software-in-the-loop simulation for Swarm Production demonstrates multiple scenarios executing multiple production orders with different product mixes. This research work also includes integrating all the systems to form a production suite. The work concludes with quantitative data acquired from the scenario-specific simulations and a formal analysis based on the results. The research contributes as a first full factory demonstration and a potential test bed for upcoming research undertakings within Swarm Production.
KW - Autonomous robots
KW - Multi-robot execution
KW - Simulation
KW - Structural self-organisation
KW - Swarm production
KW - Topology optimisation
UR - http://www.scopus.com/inward/record.url?scp=85217767796&partnerID=8YFLogxK
U2 - 10.1016/j.jmsy.2025.01.020
DO - 10.1016/j.jmsy.2025.01.020
M3 - Journal article
SN - 0278-6125
VL - 79
SP - 484
EP - 503
JO - Journal of Manufacturing Systems
JF - Journal of Manufacturing Systems
ER -