Towards Goal-Oriented Semantic Orchestration for Resource-Aware Robotic Function Offloading

Abstract— Edge computing is increasingly seen as the perfect enabler for distributed robotic applications requiring low latency and high adaptability. However, current approaches offload computational tasks exclusively to edge nodes, overlooking scenarios where onboard processing can meet performance targets. Therefore, we propose a framework that unifies offloading decision and resource orchestration into a single algorithm that compares different allocation strategies under accuracy, latency, and energy goals. Results show that this approach admits more functions while maintaining strict performance targets, offering substantial gains in overall system utility