After the spread stage, the use of IoT is moving to the next stage. Why is edge computing said to be the key to IoT evolution? Learn what edge computing is, the connection between IoT and edge computing, and where edge computing is used.
The criticality of edge computing is gaining attention
IoT is expected to become the norm in many fields in the future industry. What is edge computing that attracts attention?
What is edge computing?
In a computer network, a cloud like a data center is sometimes referred to as “central”. So where do the edges and edges, or “edges,” stand for? It may be a smartphone or tablet that can be touched by human hands, or it may be an industrial device that is active at the forefront of the field. Edge computing is the idea of processing data in a place close to the “edge”.
However, edge computing does not limit the processing of data to the edge side. Edge computing is a distributed open IT architecture, and its greatest feature is said to be distributed processing power.
While improving the response speed by processing data on the edge side, data for which speed is not required is accumulated on the cloud side. By performing distributed processing in this way, the purpose of edge computing is to maximize the value of data and optimize the utilization method required for each data
Edge computing adds accuracy to IoT
Edge computing is said to be an essential technology to take IoT to the next level. How will edge computing change the IoT?
The most obvious effect is that the real-time nature of the information is improved. Sending, processing, and receiving data to the cloud can result in a time lag of hundreds of milliseconds to seconds. Even if the sensor becomes more sophisticated and accurate data is sent, if there is a time lag in data processing, the response of the equipment will slow down and accurate operation will not be possible. Edge computing processes the required data on the edge side, reducing time lag and enabling real-time response.
Edge computing also benefits from optimizing traffic. With the spread of IoT, huge amounts of data will be transmitted and received, and the amount of data will continue to swell in the future. If you aggregate all of this into the cloud, you’ll experience data jams. By distributing processing on the edge side, data congestion can be avoided and communication costs can be reduced.
Another advantage of edge computing is that it enhances information security. Processing data on the edge side reduces communication with the outside world. This reduces exposure to external attacks and reduces the risk of leakage.
In addition, edge computing is also effective as a business continuity plan. If you’re dealing with the data you need on the edge side, you can continue to run even if the cloud goes down. As distributed processing power is a feature of edge computing, distributing data also disperses risk and increases business continuity.
The link between edge computing and IoT creates these benefits. These enhance the accuracy of data, the accuracy of equipment, and the accuracy of the enterprise, and take IoT, which is often used only for information visualization, to the next level.
Use cases of edge computing
Where is edge computing, which produces such many benefits, actually playing an active role? Here are some examples.
- Environmental sensing that detects failures with sound
A skilled technician may hear the sound of a running machine to determine if there are any signs of failure or if maintenance is needed. It is difficult to express it as an accurate indicator, but there are aspects that have supported Japanese manufacturing technology by such “on-site cans”.
It is also a challenge in the Japanese manufacturing industry to pass on such know-how on the preservation of equipment and equipment that belongs to skilled engineers through digitization.
To make this possible, solutions have been developed to detect noise and provide signs of failure. The operation sound is collected by the microphone and the abnormality degree is calculated. This data is aggregated and stored in the cloud, used for big data analysis and machine learning, and at the same time sent to the department that performs remote monitoring.
Edge computing has been introduced for processing from sound collection to anomalies. Anomalies can be detected on the edge side, enabling advanced processing with low latency.
- Real-time weather prediction simulation
Edge computing’s distributed processing power is very well suited for processing data in real time, which varies from region to region. From the possibility, weather prediction simulations were considered.
To improve the accuracy of weather predictions at a location, you need to subdivide the extent to which you make predictions. However, the more fragmented, the more data is sent to the data center. There were limits to this communication volume in terms of cost, speed, and processing capacity, and as a result, there was a limit to the subdivision of the prediction range.
But with edge computing, you can push that limit. By processing at the edge for each subdivided range and sending the data necessary for wide-area prediction to the cloud, it can be used where the necessary predictions are required.
Edge computing may soon be able to predict highly accurate, real-time weather at that point.
- Improving the efficiency of agriculture with drones and robotics
Edge computing is also expanding its potential in agriculture.
In the past, it was necessary for people to go to the site to visually determine whether the harvest time was or not. A system that changes this work into people and is performed by drones and robots has been put to practical use.
Equipped with image recognition by camera and AI, AI identifies acquired images by edge computing, analyzes ripeness from real colors and shapes, and detects pest diseases from leaf colors and shapes. Together, these can be visualized to visualize the number of harvests within the active range and predict the optimal harvest time.
Edge computing is applied to this process, and the robot can operate smoothly by separating image identification at the edge and data analysis in the cloud.
- Optimize store operations with staff behavior analysis
In a small store, such as a convenience store, a single staff member must be responsible for multiple tasks. I am also in charge of product inspection, cash register work, and, in some cases, cooking work. Therefore, it is necessary to move frequently in addition to one place in the store.
In order to streamline the movement of staff in the store, a solution has been developed to analyze staff behavior and optimize in-store placement. The camera footage placed in the store is sent to the computer in the store to identify the person at the edge and track it. As a result, data on how staff is moving is accumulated in the cloud, and the in-store arrangement that can realize the optimal movement line by data analysis is calculated.
This technology can be applied not only to staff, but also to customer purchasing behavior such as what kind of product display promotes purchasing behavior. The challenge of such use is to prevent the identification of individual identification elements for the extracted person data.
- Smart factory that controls the operation situation in real time Among the systems using
IoT and edge computing in many fields, the fastest evolution continues in the manufacturing field, especially the smart factory.
A smart factory is a smart factory where all devices and sensors are connected to a network. This not only visualizes the operation status, but also connects all kinds of information, such as forecasting and corrections on the production management side, inventory management in terms of logistics and import and exit management.
In addition, some factories have introduced systems that detect the movement of people placed on production lines by video and prevent human error by checking abnormal behavior.
Edge computing is no longer an essential technology for smart factories that require massive data processing and real-time response.
Toward an era where more accuracy is required
In this way, edge computing is used in various places in conjunction with IoT. IoT is making it possible to visualize various things at manufacturing sites, but ioT utilization is moving to the next stage. This is the second stage of ioT utilization: how to utilize the collected data. Edge computing is required to perform accurate processing without compromising the real-time nature of the data. Edge computing will become an indispensable technology for ioT utilization in the future.