What is Fog Computing?

Fog Computing is a computational model that tries to process data as close as possible closer to its source.

Processing "nodes" are created at intermediate locations between the local data sources and networks, on the one hand, and the Cloud on the other.

Unlike cloud models, Fog Computing creates a decentralised network, generating a "fog" for data processing.

What this decentralised model does is reduce the distance data must travel on a given network, thus achieving faster and less resource-intensive performance.

Differences between Thin Edge Computing and Fog Computing

To understand the differences between Thin Edge Computing and Fog Computing, again it is necessary to look at a definition of the first. Edge Computing uses the very devices that are creating the data to process and store it.

It takes the decentralised concept of Fog Computing one step further, as there are no intermediate nodes to transfer data to and the data is processed on the device itself.

Edge Computing has quickly become a catalyst for the digital transformation of the industrial sector, enabling faster deployment of industrial IoT models in the face of potential connectivity issues.

While there are some important differences between Thin Edge and Fog Computing, it is also important to stress what they have in common: both are computational models for the distribution of intelligence.

Collateral benefits of both systems include reduced latency, lower bandwidth usage (as data is filtered before being sent to the network) and improved security.

Where the data analysis takes place:

• The Edge Computing model takes data processing to the “edge”, close to the devices where the data is generated. It therefore uses the individual hardware systems, including environmental sensors, controllers, and other devices.

• The Fog Computing model takes processing one step higher in the network topology, processing data on an intermediate network located between the Cloud and the Edge. This means that Fog Computing uses a fewer nodes for data processing, and can combine data from different sources.

Technology implementation costs

Generally speaking, a Fog Computing system is more expensive to implement, as more powerful (and therefore more expensive) equipment is required.

Thin Edge Computing and Fog Computing Performance

• Edge computing provides lower latency with better security, as less data has to travel over the network.

• Fog Computing models allow data processing through devices with higher capacities, and various data sources can also be combined during the filtering process.

Fog Computing in Intelligent agriculture

There are some proposals, such as the Phenonet project (currently abandoned), for monitoring field conditions and plant growth. The project proposes the deployment of two types of nodes, sensors and gateways, although the functions could take place simultaneously in some of them. It also proposes the collection of this data by means of hot air balloons and mobile vehicles on the ground, to eliminate long-distance wireless communications.

Fog Computing in Public Transport

Imagine that you want to fit a set of environmental sensors to buses so that as they travel around the city they can collect data on pollution, gases, ambient noise, etc. Since in this example it is not necessary to have the data in real time, it would be sufficient to deploy gateway nodes at certain stops so that the data can be transferred from the bus to the gateway node when it stops there.

Fog Computing in Waste management

Urban waste collection and treatment is certainly one of the typical examples suggested for the use of IoT. Without going into too much detail, we can simplify the whole system by looking at the system participants:

• Local authorities that want to provide an efficient service as economically as possible, while complying with the terms of current contracts.

• Companies performing collection and further processing that want to automate and improve their processes as much as possible.

• Health authorities who want to be completely sure of how all waste products and waste treatment is carried out.

All these parties want to be sure that the others are performing the tasks with the highest quality and in compliance with regulations and contracts, and at the same time they want their own processes to be optimised.

Fog Computing in Water management

The management of water and its entire cycle in society is becoming increasingly important, from channelling to collection points, treatment, distribution, leakage, subsequent treatment, and even its final destination.

In each of these aspects, technology is now playing a fundamental role. A whole network of sensors is needed to monitor each of these processes, as accurately and with as much control as possible.

Fog Computing in Smart grids

Balancing electricity grids automatically to ensure supply to citizens is becoming more complicated every day, especially with the growth of renewable energies and the possibility of them introducing energy at almost any point in the grid.

Benefits of Fog Computing

1. Lower costs. If the bandwidth is lower, this will have an impact on costs. However, the initial investment for its implementation will be higher.

2. Fog computing generates greater agility in management and analysis in companies. It will enable improved productivity and faster creation of applications.

3. Improved security. As data is transferred over shorter distances, it is easier to control and protect data transfer.

‍

Edge Computing and Fog Computing as complementary technologies

As discussed above in the article on cloud systems, the Edge and Fog Computing models offer the greatest advantages when combined to take advantage of the benefits of both.

Using both technologies, it is possible to:

• Access more complex calculations and processing with Fog Computing

• Perform lower latency processing in decision making with Edge systems

We are already working on a functionality of our platform that will allow control of the hierarchy in multi-node networks in a very simple way. This will significantly simplify the deployment and execution of algorithms that require the combining of different scenarios in Fog Computing and Edge Computing models.

An example of this is the success story is the successful combination of both technologies in our project CONNECTS. CONNECTS project. In this project, we have developed a network of Intelligent Transformation Centres with Edge Computing.

As a result of the collaboration between Barbara IoT and the Ormazabal Corporate Technology Centre, agents in the energy value chain will be able to create new business models and optimise Operation, Maintenance and Planning in tasks required at the Transformer Substation.

Still under development, this project may require the installation of nodes at different hierarchical levels, thus working in both Edge Computing and Fog Computing.

Relationship between Fog Computing and the Cloud

To understand the relationship between Fog Computing and cloud-based models, it is first necessary to determine the definition of the latter.

Cloud computing involves a series of technologies that allow remote online access to different types of information (from software, to stored files and data processing).

The cloud is therefore a kind of remote server to which data is sent for processing and can be accessed remotely.

In contrast, Fog Computing-based systems avoid directly sending data generated by devices. Instead, they use processing centres close to where the data (from sensors to robots and IoT devices) is generated. This creates a secondary local network that acts on the data before sending it to the main network.

With this intermediate step of decentralisation, more immediate actions are achieved, which also require less bandwidth consumption.

It is also particularly useful for industrial computing environments where there are connection difficulties (low accessibility, congestion, insufficient networks, etc.).

However, it is important to understand that Fog Computing is not a replacement for Cloud infrastructures, but as a complement to them. This makes it possible for industrial companies to benefit from the possibilities of both computational models: the low latency of fog computing and the high capacities and versatility of cloud models.

‍