IDC recently released the report “IDC PeerScape: Best Practice Case of Big Data Management and Analysis Platform for Industrial Scenarios”, which summarized the four challenges and practice paths faced by industry users in the application process, and selected the best practice cases, and provided relevant guidance and suggestions for industry users for market reference.

The core value of the industrial big data platform is to establish the data element circulation and value mining system in the whole cycle, so as to realize the comprehensive upgrade of coverage capacity, production efficiency, data governance, enterprise management and business ecology. Industry involves structured, semi-structured, and unstructured data such as video, image, text, voice, log, and document in complex scenarios such as manufacturing, energy, and factories. There are various types of databases and uneven data quality. Currently, in most cases, there is a lack of unified data standards and management processes, and it is difficult for enterprises to build comprehensive industrial big data management ability at their own level. Therefore, the mature one-stop big data management platform of external manufacturers is needed to break through the bottom data barriers. Only by ensuring the safe and free flow of data, can the upper business management upgrade and operation efficiency be promoted. At the same time, expert experience is also playing an increasingly important role. Knowledge as a Service has become a trend, which packages business experience into standardized products and services to better solve the diversified needs of enterprises on a scale.

At present, the application of industrial big data is mainly in the form of single point discrete statistical analysis, and the operation process is difficult to trace, and there are barriers to data interaction, which makes it difficult to play the large-scale clustering effect. Most industrial enterprises are still in the stage of digital transformation and upgrading from 0 to 1. Enterprises themselves have mastered a large number of industrial knowhow, but lack the experience of integrating with big data and artificial intelligence technology to solve practical problems, such as multi-equipment management, data software integration, trend prediction, knowledge graph, equipment predictive maintenance, quality detection, etc. This requires manufacturers to enrich the big data technical architecture in the process of enterprise service, such as storage engine, analysis tool and industry model, to create a decoupled flexible and adaptive functional system for manufacturers, and through the core links of the product line, complete the whole process management of industrial data acquisition, storage, management and use. IDC predicts that 10 percent of China’s top 500 companies will deploy data and action feedback loop systems by 2027, resulting in higher returns on investment in data and content acquisition and analysis.

FBM232 P0926GW

FBM232 P0926GW

FBM232 P0926GW

Key challenges facing the market

Data chimneys and isolated islands are the main reasons that lead to enterprises unable to expand large-scale production and inefficient management. Industrial production involves ERP, MES, WMS and other related application systems, with complex data sources, various types, uneven quality and large amount of data. Customers also gradually realize the demand for unified management of data center and data, so as to build a professional data index system.

Traditional production and equipment control completely rely on expert experience, while personnel iteration and more precise management demands force enterprises to develop models to achieve more intelligent management, reduce personnel costs and energy consumption or improve product yield. Moreover, industrial enterprises need a unified platform to develop, manage, arrange, update and deploy related businesses.

Industrial scenarios involve the maintenance of a variety of hardware and software equipment, and the failure of any equipment may lead to long-term maintenance, resource waste, and greater economic losses. However, the operation and maintenance of these equipment requires high labor costs, personnel management is more complicated, and personnel with varying technical levels may not be able to detect and solve faults in time. This also does not allow for the interaction of multiple device data sources;

For large industrial manufacturers, they have rich industry experience and certain market barriers. When facing the demand of digital transformation and the competitive pressure of emerging technology enterprises, they are limited by their huge organizational system and technical capabilities, and need external manufacturers to provide integrated transformation capabilities. This includes diverse needs such as cloud services, intelligent computing, data governance, device management, model development, predictive operation and maintenance.

IDC observed that in the construction of industrial big data platform, data quality, model productization, and multi-level and batch verification are the keys to the successful practice of the project.