Focusing on IoT edge infrastructure from “OT monitoring” to “data analysis” –IDC

IDC Japan announced “2021 Domestic IoT Edge Infrastructure Survey: Intelligent Edge Usage Status” on April 8. The report is based on a survey conducted by the company in January. The survey was conducted on the management, business division managers, department managers, chiefs, and chiefs of domestic companies / organizations promoting IoT projects, and 476 organizations responded. In this survey, we investigated the usage patterns of IoT edge infrastructure and vendor selection criteria for IoT infrastructure users. IDC defines an “IoT three-layer model” consisting of an “IoT endpoint layer,” “edge computing layer,” and “cloud or data center layer” as the basic structure of IoT (Fig. 1). Based on the three-layer model of IoT, “IoT core infrastructure” and “IoT edge infrastructure” are defined, and IoT edge infrastructure is classified into “intelligent edge” and “other IoT edge infrastructure”. Intelligent edge refers to those that enable advanced compute processing such as data analysis, and other IoT edge infrastructures refer to those that do not perform data analysis, such as IoT gateways and routers. Figure 1: (Source: IDC Japan)

* Click to see an enlarged image In the same survey, 37.2% of respondents currently selected “data analysis processing” as the most important processing in IoT edge infrastructure, and selected data analysis processing three years later. The percentage of respondents will increase by 6.5 points to more than 40% of the respondents (Fig. 2). On the other hand, the ratio of selecting “OT (control system) monitoring / control” decreased by 10.7 points from the current 33% to about 20% of the respondents. “Data analysis (deep learning using AI)” will become the top item after three years, replacing the current top item “OT monitoring”. Figure 2: The most important processing performed by IoT edge infrastructure (Source: IDC Japan)

* Click to see an enlarged image. Regarding the data used in the analysis process, “image data (video)” will be the top item after 3 years, and “image data (still image)” and “system log (operation status)”. ) ”“ Position data ”followed. The top three reasons for performing data analysis processing on the IoT edge infrastructure are “considering the overall balance between the IoT core and the IoT edge infrastructure,” “fast data processing speed,” and “fast data analysis speed.” Items related to security such as “for” and “for preventing data leakage” continued. As data analysis processing in IoT edge infrastructure increases and the needs for intelligent edges expand, the number of cases where AI (artificial intelligence) is installed in intelligent edges and image data is used for data analysis will increase. In this way, IDC believes that it will be necessary to respond to the diversifying data analysis in the future in terms of processing performance and to dispel customer concerns about security. When asked about their intention to adopt the IoT edge infrastructure, which performs the most important processing, nearly half of the respondents now chose “general purpose server”. Three years later, the percentage of those who chose general-purpose servers decreased, and the percentage of those who chose “IoT edge-only products” and “cloud service vendors’ IoT edge services” increased. The ratio of selecting these IoT edge infrastructures is around 30%, which is almost in equilibrium. Masayuki Shimokawabe, IDC Japan Enterprise Infrastructure Research Manager, said, “It is important for IoT infrastructure vendors to formulate competitive strategies and alliance strategies with public cloud service vendors in addition to product strategies such as strengthening the product lineup of Intelligent Edge. Is. ”