Programmable 5G Network for Enabling Smart Port Operation
Abstract
Digitalisation and automation are key trends in maritime port operations as the industry is moving towards smarter and more sustainable future. In order to realise fully automated systems where yard cranes, autonomous guided vehicles (AGVs), and connected vessels are working in tight coordination, reliable wireless connectivity is essential to facilitate closed-loop control orchestration. Traditionally, wireless network is viewed as just a simple bit-pipes providing best efforts connectivity, which is insufficient to support autonomous port operations. However, with recent move towards cloudified 5G and increasing trend towards network API, wireless network is transformed into intelligent and programmable platform, providing opportunities to the developers and end users to exercise control and interact with the network based on their use cases and needs. This work presents how this is introduced into autonomous systems, the benefits it can bring, and the important considerations as it pertains to orchestration and optimizing this new degree of control. We first outline where relevant network control, in the form of network API, resides in the architecture. Particularly, we introduce the notion of quality-on-demand (QoD), review recent proof-of-concepts on its use case, and show how it could potentially make a difference in port automation through some illustrative examples. We then extend and pose the problem statement when we start to scale to multiple systems and show that in the extreme case whereby all systems greedily request for QoD privileges, the gains can become diminishing. We then set up a centralised, global optimisation problem over time and space that motivates having coordinated ways to schedule systems calls for QoD privileges at different time that minimises collisions or flooding, hence leading to performance improvement. Our proposed scheduling approach is practical and lightweight. It aligns well with ongoing industrial transformation, network API exposure architectures, and current maritime digitalisation trends/needs. It demonstrates that open network APIs, paired with provable online control, provide a practical foundation for safe, efficient, and scalable smart-port operations.