Will edge computing kill the cloud?
Andrew Lambert CEng FIET, CEO Electronic Media Services Ltd, will discuss the role of edge computing and edge AI and its use in 5G networks and impact on cloud computing.
The evolution from local towards cloud-based virtualised data storage, computation, network management, applications and workspaces has changed the way we use our digital services. This has brought some clear benefits over traditional systems, such as easy management, universal availability and decreased hardware requirements for deploying innovative new services. We are witnessing a change from separate person-to-person, person-to-machine and machine-to-machine (IoT) computing towards the Internet of Everything (IoE) computing. In today’s cloud systems, all data processing and decision-making logic is typically handled at data centres, which is not optimal from the viewpoint of performance, efficiency and reliability, security or privacy.
Edge computing is a key technology to unleash the full potential of 5G technology since it enables deploying computational tasks near the end-devices and therefore opens novel business opportunities around real-time analytics and actionable intelligence at the edge. It provides computational capacity near the source of the data, allowing various data pre-processing, refining and analysis functions to reduce the amount of data to be sent to cloud servers thereby reducing the load inflicted on core networks and data centres.
Many IoT devices are already embedding Machine Learning models to reduce the volume of data transferred to the cloud. There are a range of CCTV cameras that can already perform image analytics, in real-time, to extract and classify objects. However, in safety-critical applications even the delay of sending this pre-classified data to the cloud may not be acceptable. For example, in the construction sector there is a move towards using CCTV to protect exclusion zones when objects are being craned or other overhead work is being undertaken. Machine Learning embedded in a CCTV can be used to extract the location of a person entering the exclusion zone. Powerful analytics opens the possibility for the size, shape and location of this zone to be dynamic as the crane moves or the overhead work progresses. An edge computing appliance can process inputs from cloud-based project planning applications about scheduled work plans, camera data about extracted objects plus position information from the various sensors, in real-time enabling smaller exclusion zones which can improve site productivity.
One of the core pillars of 5G is a new LPWAN mode to support massive IoT (e.g. 1 million devices per square kilometre). The lower frequency (700Mhz) spectrum being made available for 5G will improve rural and remote coverage. However, this spectrum will not support high-speed low latency connections and edge computing will be required to provide local real-time analytics.
In summary, edge computing will help deliver real-time actionable intelligence at the edge for many applications. However, the cloud will still be required to provide reliability, security, scalability and macro-level analytics of the processed data from the edge.
EMS is a member of an EU CELTIC-NEXT consortium. The UK partners, funded by innovateUK, are focusing on the application of AI/ML in a distributed multi-tier network incorporating Cloud Computing, Edge/MEC devices, light-weight virtualisation technologies (NFV) and cybersecurity in a 5G environment.
To read more from #ConnectivityForAll Campaign Week visit our landing page by clicking here!