Edge Computing
What is it
Edge computing is a distributed way of moving compute and storage closer to where data is generated, so less raw data traverses the network, reducing latency and improving efficiency and responsiveness.
Where is it
In Edge computing, many data processing and filtering tasks take place in nodes or devices on the outer boundary of a computer network, or Edge. These interact intelligently with a central Cloud.
In the case of One Big Circle’s AIVR rail infrastructure monitoring platform, Edge devices include an array of sensors such as video and thermal imaging and they also perform a wide variety of onboard processing.
Why was it needed
Many computer networks must handle raw data at scale – but the movement of large volumes of data may overload the Cloud, causing latency. Private network links can resolve the problem, but at a cost that would make some applications uneconomical.
Any rail infrastructure monitoring system that relies on the use of video data at scale must overcome this difficulty. OBC had to find smart ways make the flow of data between Edge devices and the Cloud more efficient.
When was it introduced
While Microsoft says that Edge computing was explicitly conceived during discussions in October 2008 about the future of cloud computing, the underlying idea behind Edge Computing first emerged much earlier. Between the 1960s and 1980s, computing had swung between centralized mainframes and more distributed client-server systems. The rise of the web in the 1990s exposed scalability and congestion problems in purely centralized server models, especially as more users tried to access the same distant servers. In 1998, Akamai and similar content delivery networks (CDNs) proposed distributing caching servers geographically close to users to reduce congestion and latency.
With the huge increase in the movement of data brought about by the rise of the Internet of Things in the 2010s, the advantages of Edge computing became increasingly relevant through the 2020s.
OBC began to use Edge computing in its AIVR system from its inception in 2018.
How does it work
AIVR’s Edge devices are in effect into autonomous actors that cooperate with the Cloud to deliver useful data, while withholding data without utility. Even so, ultimately Edge remains subordinate to Cloud: the flow of data and information from the Edge is orchestrated by an AI-based application in the Cloud called Dynamic Demand.
READ MORE in our white paper Automation in overdrive: engineering a revolution.