Performance and Cost of Dynamic Reliability: A 6G Industrial Environment Perspective

The use of robots in modern manufacturing facilities, where also human workers operate, offers substantial advantages, but also introduces safety hazards that need to be mitigated with all possible means. Ultra-reliable communication services offered by 5G mobile networks allow for automated remote control of the machines, thus enabling a new generation of network applications that can enforce safety procedures in a hyper-connected industrial environment. However, in the context of Human-Robot Collaboration scenarios very strict safety requirements must be satisfied, determining communication reliability levels to be guaranteed that go beyond the up to 10-5 capabilities of 5G ultra-reliable deployments. Moreover, in 5G the reliability is imposed as fixed requirement to be achieved at all times, which may lead to resource overprovisioning. Resource efficiency can be improved by dynamically adapting the target network reliability level based on the actual needs of the connected devices, still guaranteeing safe human-robot interactions. The idea of dynamic reliability applied to 6G industrial environments, introduced in our previous work, is further investigated in this paper by evaluating performance and costs of the approach, and by exploring how to leverage the knowledge of the specific nature of the task performed by robots cooperating with humans. The results of our study allow to quantify the resource gain that can be achieved with dynamic reliability compared to the solution currently offered by 5G systems, as well as the cost that must be factored in in terms of network signalling overhead.