Semi-heterarchical agile control architecture with intelligent product-driven scheduling

The paper proposes a 2-layer holonic control architecture for optimizing production on both long- and short term. This architecture is intended for industrial environments affected by perturbations like: resource breakdown and recovery, exhausted part stocks, variable processing and transporting times. The proposed planning, scheduling and control models with their implementation frame are generic; the structuring of the decisional entities (Active Holon Entities – Expertise and Order) and the distributed decision making do not rely on specific or proprietary technologies. The method is based on: automatic switching between the global “batch” planning and scheduling horizon and the local “packet” scheduling horizon for resource (re)assignment – thus combining optimality, agility and fault-tolerance in business-oriented scenarios, and embedding holonic characteristics – autonomy and cooperation – in intelligent devices which assist products during their scheduling, execution and tracking – thus bringing closer the physical and decisional parts of entities composing a more performing, robust and agile control system.