A Novel Distributed Orchestration Engine for Time-Sensitive Robotic Service Orchestration Based on Cloud-Edge Collaboration

摘要

Applying service orchestration to robot application development can significantly accelerate development processes and facilitate online reconstruction. However, the centralized execution of traditional orchestration engines leads to performance and security challenges, making them inadequate for time-sensitive robot applications. To address these challenges, this article proposes a novel distributed orchestration engine based on a cloud-edge collaboration architecture and execution method that migrates orchestration rules to the edge for collaborative execution. Additionally, we propose a dynamic container deployment method based on a minimum communication latency strategy to enhance the performance of the proposed engine in executing complex orchestration tasks involving algorithm services. The proposed engine is implemented based on a two-layer data routing mechanism and compared with traditional orchestration execution through simulation experiments and application cases. The results demonstrate that it has significant advantages in execution efficiency, load pressure reduction, and data privacy.