Effective production scheduling depends on both the chosen algorithms and the system landscape, including data quality, integration, IT constraints, and operational rules. A decision model maps problem settings and requirements to algorithm classes to ensure alignment with objectives and existing systems. Organizational roles, processes, and ERP/MES/shop-floor interactions shape how schedules are interpreted and adjusted, so technical and organizational views must be integrated. Future work should add role-based, process-oriented, and explainability aspects to form a holistic framework consistent with Industry 4.0.
| Topic Fields | |
| Published | 2025 |
| Involved Institutes | |
| Project Type | ICNAP Community Study |
| Result Type | |
| Responsibles |
The purpose of the system is to guide the selection and deployment of production scheduling approaches that fit specific objectives, constraints, and existing ERP/MES/shop-floor environments. Its core functionality centers on a decision model that maps problem settings and requirement profiles—such as data quality, integration latency, uncertainty handling, and explainability—to classes of algorithms including MILP, constraint programming, heuristics, metaheuristics, simulation-based methods, and reinforcement learning. The architectural approach is modular and service-oriented, separating data ingestion, normalization, constraint modeling, algorithm selection, scheduling execution, and feedback services. Data flows from ERP master data and orders, MES events and states, and shop-floor telemetry; these are consolidated into a canonical model, enriched with operational rules and role-based decision checkpoints, then used to build scheduling instances whose outputs are returned to ERP/MES and surfaced to planners and operators.
Key technologies and standards include REST/GraphQL APIs, message queues (AMQP/MQTT), OPC UA for equipment connectivity, ISA-95/ISA-88 for manufacturing data structuring, BPMN for process and role modeling, and TLS for secure transport. The deployment model supports on-premises, private cloud, or hybrid setups, typically containerized and orchestrated (e.g., Kubernetes) with CI/CD and infrastructure-as-code. Target users are production planners, operations managers, MES/ERP administrators, and enterprise/OT architects.
Performance considerations address scheduling latency, throughput, horizon size, and responsiveness to real-time events; security covers role-based access control, audit logging, encryption in transit and at rest, and network segmentation between IT and OT. Notable constraints include variable data quality, integration limitations, compute bounds for exact methods, and requirements for human-in-the-loop oversight and explainability. Scalability is achieved through horizontal scaling of stateless services, distributed solve farms, and partitioned scheduling by area or product family. External integrations include ERP (e.g., SAP), MES, CMMS, PLM, APS, BI/analytics platforms, and shop-floor devices via OPC UA and industrial gateways.
Contact us to get in touch! With a membership, you’ll gain full access to all project information and updates.
© Fraunhofer 2026
