▲AIoT concept, Korea Intelligence of Things Association


AI+IoT: Massive Data Generation and Optimization
Edge computing requires hardware and network infrastructure.

[Editor's Note] We often see words with the word "smart" in them. For example, LG Electronics' home appliances, which I've been keeping an eye on lately, can be controlled with a single ThinQ app. Kimchi refrigerators use artificial intelligence to customize kimchi storage, and air purifiers automatically monitor air quality and operate accordingly. The current trend is evolving from simply "obedient devices" to "devices that do things for themselves." Beyond AI and IoT, the two technologies are now being integrated, and this is being applied not only to home appliances but also to various other fields, proving our increasingly convenient lives. In this article, we briefly covered the basic concepts and trends of AIoT.

■ What is AIoT?

AI (Artificial Intelligence) is a computing technology that learns and solves problems like the human brain, and IoT (Internet of Things) is a technology that connects objects equipped with sensors through internet communication functions.

AIoT (AI of Things, Artificial Intelligence of Things) is a combination of artificial intelligence (AI) and the Internet of Things (IoT). It is a technology that creates algorithms that classify, analyze, and predict large-scale data collected from various fields through IoT using AI's thinking, learning, and self-development with human intelligence.

Ultimately, AIoT technology is not simply about device interaction, but rather about building algorithms that optimize the system using acquired data through AI.

■ Why AIoT?

Because AI and IoT are complementary to each other, their combination is inevitable.

IoT devices are interconnected, communicate, and transmit the generated data to the cloud for processing. However, the cloud was difficult to scale in proportion to the massive data volume, which eventually led to the problem of cloud overload.

Furthermore, traffic delays and congestion in areas such as autonomous vehicles that require faster decision-making and processing compared to the increasing amount of data can be fatal.

Ultimately, a solution was developed to process the massive amount of data provided by IoT devices at ultra-high speeds at the edge or on the devices themselves, and this led to the concept of 'edge computing'.

■ AIoT and Edge Computing

'Edge computing' is a technology that emerged alongside AIoT, and it creates an efficient environment by performing processing previously performed in the cloud, such as learning and inference, at the edge or on the device itself.

Edge computing can reduce latency, improve reliability, and reduce costs.

Edge computing requires the support of advanced hardware technology. The integration of AI semiconductors into IoT devices allows for advanced sensing, enabling enhanced data collection and analysis.

Additionally, effective AIoT systems require a robust network to support them. As 5G communications environments evolve, edge computing delivers speed, reliability, low latency, and increased capacity to support thousands of data-generating IoT devices.

■ Existing AIoT cases

AIoT has limitless potential. It can be applied anywhere that requires information optimization through data analysis. In particular, it is transforming the landscape of various industries, including manufacturing, automobiles, telecommunications, and aerospace.

The Fourth Industrial Revolution is driving a digital transformation for manufacturing companies, driving the creation of smart factories. By empowering AI-trained computers to solve problems and make decisions using the vast network data generated by IoT-enabled equipment, we can move beyond simple industrial automation and achieve intelligent factories.

Self-driving cars are the epitome of AIoT. Self-driving cars, equipped with AIoT technology, rapidly collect and analyze data through sensors. Based on external factors such as road conditions and obstacles, as well as the driver's status and behavior, they can make decisions and initiate or halt driving.

AIoT applications are becoming increasingly common, not only in industry but also in everyday life. As mentioned earlier, AIoT technology can be put to practical use in addressing social issues, such as controlling home appliances with a single app in the smart home, and notifying users of life hazards, epidemics, and hazardous substances.

Recently, medical robots developed in the medical field have been performing tasks such as disinfecting facilities, sorting recyclables, and transporting materials within hospitals. AIoT can be used to help society at large through remote diagnosis and care, and the identification and treatment of critical patients.

■ Future AIoT market

According to Research and Markets, the AIoT market is expected to grow to over $65 billion by 2025.

On the 21st, the Ministry of Science and ICT announced that it would establish a core foundation for leading the digital transformation by discovering killer services that apply new technologies of the Intelligent Internet of Things (AIoT).

In addition, it was announced that a total of 8.55 billion won will be provided to 12 projects in fields such as smart home, disaster safety, agriculture, forestry, livestock, and fisheries to discover AIoT projects with high public perception performance this year.

AIoT, which combines the benefits of AI and IoT, requires the right hardware, software, network, and system design. Later, we will delve into the technical aspects of AIoT and explore notable companies.