Thin Edge Computing vs Thick Edge Computing

The growth of Edge technology is causing a revolution in the transition to Industry 4.0 and in turn the need arises to distinguish different parts of the EDGE with different functionalities.

Technology

Concepts such as thin edge and thick edge arise from the exponential growth of a technology, Edge Computing, which all industry experts highlight as key to the digital transformation of critical industrial infrastructures.

As new use cases emerge, edge diversification continues to expand along with technology and business requirements.

To better understand these concepts, we want to first understand what they mean , then compare the two extremes and finally analyse some of their technical requirements for industrial implementations.

Download the Industrial Edge Computing Barometer

Thick Edge or Thin Edge - what's the difference?

The distinction between a thin edge and thick edge use case comes down to where data is processed and intelligence is generated.

So how do we know what to use and where to use it to improve our industrial process? This is one of the main debates within development teams: do your devices need thick edge, with highly analytical processing power, or thin edge with lighter and therefore less expensive computing?

Industry 4.0

What is Thick Edge?

When data processing is performed at the point closest to the network and furthest away from the devices, we speak of "Thick-Edge". This occurs at a distance of between 100 m to 40 km from the devices, and is carried out by very high power Edge Nodes, or in some cases even embedded in the core network equipment itself. This is the case for example with some 5G communication towers, which can perform data storage and processing while avoiding unnecessary latency when the communicating devices are on the same network.

Thick edge devices can perform extremely intensive analytics for a complete connected asset environment. They are optimal for situations where fully self-contained, high-performance analysis and operational visualisation is desired and where size, power and the addition of a new physical component are not an issue.

But there are many situations where size, cost and power are important constraints.

This is where thin edge technology comes in.

Recommended reading: IoT Edge Computing, Edge Nodes and use cases in the Industrial Sector

What is Thin Edge?

If the data processing is carried out on network equipment or data aggregators located in the local network itself, we speak of "Thin Edge". The physical distances in these cases can range from 1 to 100 metres, and it is characterised by being carried out by medium-powered Edge Nodes, 1GHz and no more than 8GB of RAM, which in many cases also act as data concentrators, IoT gateways, or even intelligent industrial automation equipment.

Thin edge implementations are typically much lighter than thick edge in terms of physical size, product cost, power requirements and installation complexity. They also operate with minimal resources on small, lightweight computing devices , such as modern network routers, set-top boxes or display units.

Thin edge devices until now could perform less intensive analysis on a single or small number of connected assets, but did so semi-autonomously. If a problem was detected, the device would "call for help", alerting online platforms to vibration or moisture measurements, for example, outside locally manageable thresholds.

Recently, there has been a clear trend to integrate more intelligence in smaller devices, thanks to platforms such as Barbara, capable of enabling AI applications on these devices as well.

Graph of different types of edge computing thin edge and thick edge

Lightweight, low-power perimeter devices for harsh, low-connectivity work environments are key to automating and improving processes, as well as enabling new business avenues.

In many industries, such as IIoT, these devices are located far away from where data is processed. Therefore, any data coming from these remote devices is usually delayed and irregular. In some high-impact or high-cost service industries this wait is critical.

Thick Edge vs. Thing Edge

The solution? Thin edge computing that enables real-time processing of usage data: analysing data closer to the source, reducing latency and saving costs of moving and storing it in the cloud.

Industries need to be able to extract even more value from their products through increased uptime, additional operational intelligence and improved user experience and have seen the edge as a perfect opportunity to achieve this.

But there are many nuances of edge and they all coexist for provide intelligence on remote assets individual or groups of co-located assets, as needed.

Barbara Edge Platform

While an increasing number of IoT use cases demand a greater degree of edge processing, solutions at the edge are still grappling with the challenges of secure connectivity and application management. This is where Barbara comes in: our platform for edge nodes conceived with security by design that enables deploy, manage and configure applications centrally on the nodes in a single click.

Through the Barbara Edge Platform we can deploy artificial intelligence applications from a centralised point. These applications can be protection and control algorithms, monitoring algorithms, energy balance algorithms, etc... In addition, in each edge node we can deploy up to 5 different applications from different authors and we are able to communicate one node with another so that we avoid centralised infrastructures with higher costs and security risks.

We are moving from a model where the intelligence was in the industrial equipment, in the hardware, to a model where the intelligence resides in the software.

Edge computing follows a clear trend in industrial digitisation aimed at providing intelligence to distributed elements, and it is essential to have specialised providers such as Barbara.


If you are interested in this article, contact us to learn more about Barbara IoT Technology and request a personalised demo.