Edge computing, what it is and how it works
10 June 2020
We often speak about Edge computing, Industry 4.0, Smart Processes, Smart Agricolture and Smart Manufacturing …. but what do they really mean? These terms often indicate the digital technologies used in many industries and different sectors to improve efficiency and productivity, while improving products and processes.
The issue of Edge computing is becoming increasingly important among emerging technologies, connected to IoT (and its evolutions such as the IoE – internet of everything and the IIoT -Industrial IoT), and to 5G. It concerns the so-called fourth industrial revolution (I4.0) and also the global and pervasive digitalization. In this way, the opinion of Peter Levine (of Andreessen Horowitz) could be considered as an extreme, since he foresees that edge computing will soon declare the end of cloud computing. Moreover, Gartner confirms the growing and relevant interest in the issue by estimating that by 2022, the 75% of the data generated by companies will be processed outside the traditional data center, by technologies such as intelligent edge computing.
There are many definitions of Edge computing online, in our opinion, if we consider that edge computing is implemented essentially through the use of devices named “edge computers”, we could say that it is a “way” to run applications on local computational systems that process data directly in the field, freeing them from close dependence with remote data centers, thus reducing Cloud congestion. The proliferation of connected devices, together with the possibility of processing “on site” and in real time (or ” near-real-time “), invite the community of system integrators, designers, systems engineers and user companies to adopt solutions that use edge computers where to run edge analytics services.
What are the differences between Edge Computing and Cloud Computing?
In our opinion, edge computing is at the same time a bridge and an alternative transit between the field levels and the cloud levels: if, on the one hand, the main focus, even if not exclusive, of the IoT is the level of field, sensors on objects and people, and on the other the Cloud is the natural location where the potential of the so-called “Big Data” technologies are fully expressed, edge computing is the enabling technology that allows you to implement, “on the field ”, some of the features typically destined for the cloud but which, if they were relegated exclusively to the cloud, would be limiting. In fact, some of the advantages obtainable with an Edge Computing solution in the IoT field, can be expressed in terms of:
– Resilience, especially with regard to connectivity problems and secondly also with power supply problems: continuity of power supply and data network are difficult to guarantee 100%; the absence of power supply and connectivity in the field must always be taken into account, especially for “safety critical” applications, both when the causes are physiological, and when they are malicious, negligent or even intentional; it is therefore essential, in different types of applications, to have “system logics”, algorithms, which are able to keep the basic functions of an IoT system active even when there is no possibility of connecting to the Cloud;
– Timeliness of the action, both when it comes to generating alarms and when it comes to real feedback typical of control systems, meaning the term control in the engineering sense; it is the so-called “real time” or “almost real time”, ie the low “latency” of the system; It is therefore essential, in some types of applications, to have the ability to generate alarms or feedbacks quickly, lower than those typically guaranteed with the cloud (and this even where it is often thought – erroneously – that 5G is in itself the solution, because in any case the low latency envisaged by 5G remains influenced by the unpredictable coexistence of many applications that divide the available bandwidth).
Edge computing is therefore an increasingly relevant topic and the arguments set out above denote a level of strategic nature of this technology. Not to be overlooked, among the features of an edge computer, is not only the possibility of “running” algorithms, also of the “AI – Artificial Intelligence” type (or at least ML, Machine learning), but also the ability to simultaneously support a high number of communication protocols in order to expand the range of connectable devices and therefore the universality of the solution (expressed in terms of flexibility, sum of modularity and scalability).
This is our vision that we try to transpose into our technological products, the result of planning and concrete skills, which combine highly technological know-how with the quality and safety of “made in Italy”.