Edge computing is a solution that prioritizes data processing for low-latency and specialised applications, particularly those involving complex and mission-critical processes. An ideal example of an edge device is a ship, which heavily relies on its onboard computing capabilities to autonomously perform various navigation tasks. The transportation sector, including shipping, has witnessed the emergence of software-defined vehicles where sensor data, analytics, and algorithms play a pivotal role in optimising operations such as propulsion, cargo handling, energy management, communication, and human-machine interactions. This chapter delves into the significance of edge computing in the shipping industry, outlining its ability to enhance operational efficiencies. It explores the specific user requirements for an effective edge computing solution and highlights the role of AI in enabling scalable and continuous computation. Additionally, the chapter provides a comprehensive overview of edge infrastructure, platform requirements, and considerations pertaining to data and AI at the edge. The discussion incorporates the Mayflower autonomous ship as a case study, illustrating the criticality of edge computing for the future of shipping.
TopEdge Computing Overview
Edge computing minimises latency by bringing computational workloads closer to the data source and the location where actions need to be taken – removing the distance between collecting data, processing it, and making those results available for use in decision making. Data is generated from diverse sources and at different scales, encompassing equipment, devices, individuals, and processes.
The concept of edge computing is closely intertwined with cloud-native development and decisions regarding where computational tasks should be executed. It prompts considerations on which data should be transmitted back to the cloud for further processing.
Telecommunication providers often refer to the network edge, which offers a chance to deploy edge infrastructure and leverage communication capabilities, particularly with technologies like 5G. This presents an opportunity to enhance edge capacity and optimize the utilization of communication resources.
The integration of compute capabilities into modern industrial equipment has opened up new possibilities for performing a wide range of tasks. For instance, cameras on assembly machines can be utilized for quality control analytics. Production processes can be optimised, comprehensive management of distribution centres becomes feasible, and software containers can be deployed in a suitable runtime environment enabling elastic scalability.
However, managing these deployments poses significant challenges. With an estimated 15 billion devices worldwide, enterprises must oversee the provisioning, deployment and ongoing management of thousands of devices, each with different specifications, purposes, operating systems, and workloads. While the provisioning process can be done in a controlled environment to ensure full compliance, security becomes a major concern as these devices operate outside the traditional confines of IT data centres. They lack the typical safeguards found in hybrid cloud environments, such as physical barriers and the uniformity required for certification. Ensuring the integrity and preventing tampering of edge devices becomes paramount.
To support the growing role of edge computing, it is crucial to build an ecosystem that facilitates its development. The impact of edge computing on enterprise computing is expected to be substantial, comparable to how mobile technology has transformed the consumer sector.