Edge computing- technology surrounded by questions and doubts. Will edge computing replace cloud? Is it a more sustainable solution? What are the driving factors behind its success? What edge computing and food delivery have in common? Learn this and more in the new Grow@IT Interview.

Over the last years, the Internet of Things (IoT) seemed to have passed all the thorns of Gartner’s hype cycle and finally established itself on the market. Not only the number of enterprises using and integrating IoT in operations is increasing but also IoT is steadily being incorporated into our daily lives. This includes smart homes, fitness watches, self-driving cars and many more. If we zoom in, we see the Industrial Internet of Things, which refers to interconnected devices and instruments for industrial purposes. Despite their diverse application, all these “smart objects” have one thing in common. They generate an enormous flow of data. This amount of data should be stored and processed. This can be done using cloud computing technologies (technologies that allow data to be stored and to be processed remotely, in cloud).

However, with the increasing number of “connected devices” and their data generation, the bandwidth capacity of existing channels may not be sufficient. So, even though, in the last years, companies were actively deploying cloud computing technologies, the situation has changed, when edge computing came into play. Edge computing or “last mile” as it is sometimes described, processes and analyzes data on the source of data generation or near it. Edge computing is like food delivery. Think about a food delivery restaurant, that opens a small branch in the neighborhood. The branch, located closer to customers, delivers food faster. Orders processed at the main location would get cold long before they even reach the target place. In that case, the main restaurant is the cloud, while a branch is the edge.

To provide its employees with a deeper understanding of edge computing, as part of the internal learning and people development initiative “Grow@IT”, Siemens introduced a Learning Journey dedicated to edge computing. This Learning Journey provides an understanding of edge computing and its relevance for Siemens. It is suitable for different levels of proficiency, including Beginner, Edge App Developer, and Edge Solution Partner.

There are indeed many aspects to learn in this area, as edge computing’s popularity is rising. We turned to Franz Delcuve (Solution Architect) who agreed to clarify all our doubts about edge computing and its further development in the world of technology.

Editorial Office: While companies explore the potential of edge computing to reduce the load on data centers, we would like to hit the “last mile” and have a better understanding of it. Before talking about more “edgy” details though, we would like to know from you, Franz, how did you get involved in the topic of edge computing and what role does it play within Siemens?

Franz: I started my journey with Industrial Internet of Things and edge computing, after completing the two-year Trainee@IT program. At that time, I didn’t know much about the whole spectrum of its industrial application, only about its usage in the private sector. Nevertheless, I was eager to get involved in IoT and Edge projects at our factories.

So, I dived into the topic right away. I was analyzing the structures and needs of our factories, I was talking to the colleagues and, of course, was trying to capture the voice of the customer. Aside from this field research, I was working on the AI project aimed to reduce test efforts on the production line. The practical experience helped me to understand how to transfer the theoretical knowledge from PowerPoint slides into an operational working setup.

Speaking about the role of edge computing within Siemens, we’ve created an IoT DevOps service that is managed in close collaboration with our customers. It helps to enrich our portfolio of edge computing use cases and move factories from proof-of-concept to productive use cases.

Editorial Office: You’ve mentioned, you didn’t think about the wide spectrum of edge computing applications. Neither had some companies before. In your opinion, what factors have determined the acceleration of the adoption of edge computing by companies, and how will it affect future work?

Franz: Imagine having one place, one cloud, where all the data could be available. You could have all the necessary information right at your fingertips. Sounds perfect, except that this “cloud-first” approach does not work. Firstly, it creates a huge amount of traffic, thereby generating huge costs. Secondly, sending every bit of data to the cloud can give hackers some clues about production processes. Thus, costs and risks are some of the factors that play in favor of edge computing.

Another big driver is human error. In terms of the production process, edge computing can be used to significantly improve overall quality control and make predictive maintenance more precise. This means that processes can be adjusted in real-time, without letting the machinery downtime and excessive costs happen.

Summing up, the edge will affect future work by increasing the efficiency of production processes by reducing the margin and the occurrence of human errors.

Editorial Office: “Cloud-first” doesn’t work. So, is it Edge over Cloud? Rumors are going around, that cloud will be substituted by edge computing, that there is more potentiality in it. Is it so?

Franz: I would say that it is important to understand the essence of both offerings and your needs. Afterwards, it becomes clear that these are non-interchangeable technologies. In the world of IoT edge computing plays the role of connectivity layer, setting a connection from the production machinery into the internet as well as computing data before sending it somewhere.

On the other hand, there is the cloud level. The cloud acts as a huge data center, which is available from everywhere. Only, you do not have to use the cloud solely as a data center. You can run apps directly within the cloud to get a visualization of data, analyze it, use machine learning, or even create AI models.

In a nutshell, edge and cloud are both parts of the IoT, offering different benefits. Therefore, you need to define your requirements. Do you need low latency? What kind of data do you want to send into the cloud? To harness the potential of both technologies, I would go for a hybrid approach. For instance, use edge computing for data aggregation and filtering, but integrate the cloud for data storage and training AI models.

Editorial Office: Nowadays, however, it is also important to consider the environmental aspect in the decision-making process. For instance, cloud data centers are a significant driver of carbon emissions. What about edge computing? Is it a more sustainable solution?

Franz: Edge computing aims at reducing internet traffic and the amount of stored data within the cloud. Therefore, one can argue that edge computing might be a more sustainable solution. However, you still need to run either Industrial PCs or servers in your factory. This calls into question the sustainability of the edge.

From my point of view, a lot depends on the energy source. For instance, if you are running coal-burning power plants to generate the energy for cloud or edge operations, neither is a sustainable solution. You might create a smaller footprint by running one setup, but the overall problem remains!

The same thoughts apply to the materials supply chain: if the production of the required hardware is not sustainable, then the whole use case scenario cannot become sustainable. You need to address the problem as a whole, not just one part of it.

Editorial Office: As someone who also had to learn new things about edge computing in the beginning, do you think the Grow@IT Learning Journey “Edge Computing” is a good start for beginners as well as a useful supplement for those who want to expand their knowledge?

Franz: In my opinion, the Edge Computing Learning Journey is a good start for beginners and for those who are interested in this topic. In particular, I would highlight the videos introducing different projects and approaches not only within Siemens but also in cooperation with our customers. This gives a good overview of edge computing possibilities, also for those who are already running edge computing use cases.

Moreover, I would really recommend “Container Security Training: Security in the lifecycle of a container”. There are many pitfalls in running a secure container, and usually, the easy way to set up your container is not the secure way. This training creates awareness of the pitfalls and offers a solution on how to avoid them.

Editorial Office: Thank you, Franz, for sharing with us your thoughts and broadening our horizon on the topic of edge! Siemens employees, who are interested to learn even more about edge computing, can find the learning journey via the Grow@IT Channel.