bmta.co.uk
LEVERAGING THE EDGE IN IIoT
By Daniel Povey
Research Scientist
National Physical Laboratory (NPL)
In the age of IoT, “edge” computing is about reducing the load and reliance on cloud platforms for data storage and
processing, conferring a number of potential advantages. The “edge” is typically used to refer to a location that is on
the same premises as the user’s data-generating process.
In Industrial IoT (IIoT), it commonly refers to a point of contact in the transmission of data that is on site with the data acquisition
devices. Consequently, a processing unit at the edge is referred to as an “edge node”, which is usually just a computer.
Consider an industrial manufacturing process equipped with appropriate IIoT devices, including network-enabled
microcontrollers. Microcontrollers can operate sensors such as accelerometers, platinum resistance thermometers (PRTs),
light-level detectors, etc.
Each IIoT device (microcontroller-sensor combination) captures data at regular intervals and sends this data to a ‘somewhere’ to
be processed and eventually analysed. In a typical IoT application this ‘somewhere’ is usually a server or data centre that forms
part of a cloud platform, whereas an application that makes use of the edge will route its data through an edge node before it
reaches the cloud.