The role of edge computing in data traffic management

Edge Computing is one of the most rapidly spreading architecture in the Industrial Automation, although it is not mentioned among the Enabling Technologies of Industry 4.0. In fact this solution allows the effective implementation of other technologies in the productive environment.

The traditional control architecture of automated lines includes intelligent units - generally PLCs - acquiring data (directly or through fieldbus) and managing motors or other actuators.

Even though these units have evolved over time, they have a limit of resources. Infact, they are highly specialized for a task: the control of the production line.

The focus on data as a value, a source of knowledge, is one of the most important megatrends of recent years. Data are a necessary basis to run complex algorithms, Artificial Intelligence and Machine Learning applications.

During the regular operation of a plant it is possible to extract lots of data directly from the sensors or from the PLC that controls them. To manage them, memory and computing power are required to be allocated next to the data sources: here comes the Edge Computing.

Infact, Edge relates to the edge of the line where the device is installed. In particular, it's installed in the electrical panels, generally next to the PLCs and to the terminal blocks connected to the sensors.

In this way the data traffic is local e does not consume bandwidth in the higher level network infrastructure. It allows to minimize the transfer delay.

Like any new solution, it is important to understand its benefits in order to smartly introduce it in the production line. A real return on investment is expected.

The companies that want to understand the real advantage of this technology - as complex as useful - must choose a partner who has gained experience in many industrial sectors, such as Adgenera.

The advantages

Software Development Cost

The evolution of a plant often includes the need for new control functions to be more complex and advanced.

In some cases these functions can be integrated directly into the control units (PLC), however their development on an Edge platform is much faster and easier. Infact, the Edge platform takes advantage of advanced programming languages, availables libraries and a greater support for the debug and validation phases.

Furthermore, not intervening directly on the PLC can be essential in case of already operational systems who can’t tolerate the downtime to update the software.

The Edge Computer can "relieve" the control unit from those processes that don’t require too rigid real-time design levels and involve a large amount of data. Because of that, optimizing the size of the components is possible in the design phase.

Predictive Maintenance

A lot of data is needed to reliably predict a machine downtime or degradation. Complex algorithms are involved.

In order to get the best results, very long data sequences are often required. They will take up a large amount of memory covering months of operations.

Control platforms, such as PLCs, are not the right ones for this type of processing: they lack in resources and other necessary IT tools, such as databases, high-level languages and libraries.

The Edge Computer can effectively detect the so-called weak signals by complex processing or through artificial intelligence, in particular using PLC data. They will be a resource to return the alerts into a visualization system or a reporting one.

Although some applications can be implemented at SCADA / MES level, the Edge Computer is suggested as the most convenient platform to perform the task, because of the large amounts of data.

Machine to Machine

M2M communication is the main tool to increase the plant flexibility and the production resilience. In fact, it allows to dynamically move the workload and the information of the single production step. They are moved in a fluid way, directly from one machine to another. Generally, different machines produced by various OEMs, with distinct technologies, cannot communicate natively. On the other hand, opening a communication channel between them can be technically complex and expensive. The Edge platform is the most ideal component to achieve this kind of communication. In fact, it is able to host the protocols translation logics. Protocols adapt to the language each machine adopts and to its data speed.

Furthermore, the Edge platform can create a temporary data buffer in order to synchronize the exchange or to save transmitted data. This happens when one of the two peers - machines - is temporarily not reachable on network.

Data Gateway

Edge Computers have both typical characteristics of a PC and industrial interfaces. So, they can easily integrate into a computer network and connect to most controllers.

Therefore, through an Edge Computer it is possible to put in communication the IT infrastructure even with the simplest or most dated controls that can’t exchange data across a network. This step is possible thanks to more advanced processes, such as a shared storage.

In some cases the Edge platform itself connects directly to the new sensors without these signals being managed by the PLC. The new information becames available to the IT infrastructure.

For example, a MES could receive real-time data of the processing products along an automated line from the Edge computer. This one presides over the tracking functionality through barcode and / or RFID readers. This, without any intervention on the line control unit.

Regulatory Compliance

In specific sectors, strict regulations place important constraints on the physical location and on transmission of production data.

Specifically, sometimes critical data are not allowed to cross the physical boundary of the production area. Or again, they are not expected to be stored in a Cloud storage located in other geographical areas than the production site.

About those data coming directly from the production lines, the Edge Computer is the best tool to implement their policies on storage, processing and treatment. Only data that can be transmitted outside the production site will be routed to the IT infrastructure. The other ones will be processed locally, where they were produced.

Thanks to the successfully completed projects involving this technology, Adgenera is able to thoroughly analyze each application. Then, we design and implement the solution that best suits the context.