Abstract
Modern manufacturing systems increasingly demand flexible and interoperable automation solutions capable of integrating heterogeneous and legacy devices. This requirement is particularly relevant for small and medium-sized businesses and research environments, where vendor lock-in, system rigidity, and high demand for integration resources often hinder the adoption of advanced automation. This paper presents a vendor-independent distributed control architecture for a vision-based robotic sorting cell, designed to support modularity, adaptability, and eventual device replacement. The proposed approach models the robotic cell as a cyber-physical system composed of loosely coupled control nodes coordinated through a supervisory control layer. An event-driven asynchronous communication strategy enables interoperability among heterogeneous sensing and actuation subsystems without relying on proprietary automation platforms. The architecture has been instantiated on a laboratory-scale robotic cell integrating legacy robots, feeding systems, and multi-vendor vision devices. Late-breaking results highlight how the proposed control strategy supports flexible system evolution and simplifies integration, while trading strict quasi real-time guarantees for adaptability and scalability. The presented work contributes an architectural and control-oriented perspective on distributed automation systems, providing a foundation for future developments toward intelligent supervision, fault tolerance, and digital twin-enabled control.