Illuminating the Future Factory: The "Zhongyun" Power of Fault Prediction and Health Management

More than two thousand years ago, the "Huangdi Neijing" proposed the concept of "the best doctors treat before illness, the middle doctors treat impending illness, and the inferior doctors treat existing illness," revealing that the highest level of medical skill is to prevent disease before it occurs. This wisdom is equally applicable to modern equipment management.

In manufacturing, equipment is the core of production, and its operating condition directly affects efficiency, quality, and cost. Traditional post-maintenance often leads to unplanned downtime, resulting in production delays and safety hazards. While preventive maintenance can reduce failure rates, regular maintenance can lead to over-maintenance, spare parts backlog, increased costs, and potential downtime due to insufficient maintenance. Data from the International Automation Association shows that global manufacturers lose approximately $647 billion annually due to equipment downtime.

Predictive Health Management (PHM) As a new generation of equipment maintenance management technology, it can help enterprises break the limitations of traditional maintenance models and achieve the effect of "treating before illness". In recent years, with the rapid development and application of technologies such as sensors, artificial intelligence, industrial Internet of Things, and big data, PHM has received widespread attention and emphasis from the industry and is hailed as the "light of the future factory".

The "14th Five-Year Plan" for the Development of Intelligent Manufacturing clearly identifies equipment fault diagnosis and predictive maintenance as one of the key core technologies in the intelligent manufacturing technology research and development action. The "2024 Work Plan" of the Industrial Internet Special Working Group issued by the Ministry of Industry and Information Technology also requires the promotion of pilot construction and promotion of equipment integrity management and predictive maintenance scenarios.

PHM: Making Equipment Maintenance "Proactive"

PHM is a new maintenance support technology developed on the basis of traditional condition monitoring and fault diagnosis technologies, covering two aspects: fault prediction and health management.

Fault prediction, namely the P (prognostics) part of PHM , refers to diagnosing and predicting the current and future health status, performance degradation, and occurrence of failures based on the system's historical and current monitoring data; Health management, namely the HM (health management) part of PHM , refers to making appropriate plans, decisions, schedules, and coordination for tasks, maintenance, and support activities based on the results of diagnosis, assessment, and prediction, combined with available maintenance resources and equipment usage requirements.

The working principle of PHM is : Based on the phenomenon that most complex equipment or its components will experience performance degradation with increasing usage time. There is a time interval between the early signs of failure and the actual occurrence of the failure, which is the Remaining Useful Life (RUL). By predicting the RUL, enterprises can determine the optimal maintenance time and plan before the failure occurs; on the other hand, by analyzing the mapping relationship between failure signs and failure causes/locations, the cause and location of the failure can be predicted.

Health Degradation Curve of a Certain Product

In general, PHM is a comprehensive technology with functions such as condition monitoring, fault diagnosis, health prediction, auxiliary decision-making, and resource management. It collects equipment data and uses data analysis and modeling techniques to achieve fault warning, preventive maintenance, and provide decision support for point inspections and equipment management, thereby effectively reducing failures, lowering maintenance costs, and improving equipment management efficiency.

Currently, PHM has become a key technology for industrial enterprises to achieve remote maintenance and ensure the reliability of key equipment, and an important means of implementing intelligent factory construction. A number of professional providers and solutions focusing on PHM are gradually emerging. Wuhan Zhongyun Kangchong Technology Co., Ltd. (hereinafter referred to as "Zhongyun Technology") As a high-tech enterprise in China that has been engaged in PHM research and application earlier, relying on its strong R&D capabilities and profound industry experience, it has established intelligent maintenance capabilities for more than 1,000 enterprises, becoming one of the leading technology suppliers in the industry. Especially in the building materials industry, Zhongyun Technology achieved the number one market share in China for fault prediction and health management systems in 2020, and deeply participated in the formulation of four national standards and two industry standards, including "Intelligent Service Predictive Maintenance".

Building a Comprehensive Intelligent Maintenance System, Acting as a "Specialist Doctor" for Equipment Safe Operation

As a comprehensive technology integrating knowledge from multiple disciplines, PHM has broad application prospects but is also one of the most difficult technologies to implement, facing numerous difficulties and challenges, such as:

• 1. To promote PHM, data collection is the foundation, but currently many system development bottlenecks lie in data collection, or the lack of a complete data acquisition system, or a lack of data accumulation mechanism, or incompatibility of equipment interfaces, etc.;
• 2. The lack of fault data makes it difficult for many enterprises to build effective fault diagnosis models, and factors such as the complexity, diversity of equipment, and environmental changes also pose great challenges to enterprises in establishing reliable prediction models;
• 3. The output of the application effect of PHM projects depends on the support of industry knowledge, data application capabilities, and expert experience, which is not easy for many enterprises...

Based on a deep understanding of PHM and precise analysis of the pain points of enterprises, Zhongyun Technology has built a complete intelligent maintenance product system from hardware to software, from online to offline, comprehensively ensuring the stable operation and efficient maintenance of industrial equipment.

• More Than 100 Diagnostic Algorithm Models, Leading a New Height in Fault Diagnosis

Fault diagnosis is a core link in PHM and also the main direction of Zhongyun Technology's efforts. Relying on its self-developed fault diagnosis algorithm software, Zhongyun Technology has created more than 100 diagnostic algorithm models targeting the component level, covering mechanism models, AI models, fixed threshold models, dynamic threshold models, and trend threshold models. It is understood that These algorithm models were developed in cooperation with renowned university laboratories. Using engineering site data as the research carrier, and fully combining expert knowledge and AI algorithms, the developed models have strong versatility and accuracy, covering more than 90% of industrial equipment failures and failure modes, including sliding bearings, diesel generator systems, overhead cranes, vibrating screens, cement kilns, power transformers, and hydropower generators.

To further improve the applicability and accuracy of the algorithm models, Zhongyun Technology has set up large-category models and further subdivided these large-category models. Enterprises can specify data to enter the corresponding models according to their own equipment characteristics and needs, achieving fast and efficient model creation and training. In addition, Zhongyun Technology has integrated various data analysis methods such as time-domain waveforms and frequency-domain waveforms, which can maximize the acquisition of equipment status information, extract fault characteristic values independent of equipment parameters and operating conditions, effectively solving the problem of small sample data faced by enterprises, and reducing the impact of equipment operating conditions on fault models, improving model accuracy. Currently, Zhongyun's diagnostic algorithm models have been applied in numerous enterprises, with an intelligent diagnosis accuracy rate of up to 95% for equipment failures and a 0% equipment failure miss rate.

In addition, these algorithm models are highly flexible and scalable, allowing for customization and optimization based on the characteristics and operating environment of enterprise equipment. They also support integration with application ends and can be easily deployed into enterprise production environments.

• IEPlat Tengyun and DeepSeek join forces to lead the upgrade of equipment maintenance management

Under the wave of digital transformation, enterprises are increasingly demanding intelligent and integrated equipment maintenance management. Zhongyun's new generation of intelligent maintenance platform—IEPlat Tengyun ，taking equipment as the core, integrates cutting-edge technologies such as the Internet of Things, big data, and artificial intelligence, providing integrated functions such as data acquisition, real-time monitoring, remote care, intelligent early warning, fault diagnosis, and equipment management; and integrates the DeepSeek algorithm, capable of processing multi-source heterogeneous data in real time and, through dynamic knowledge distillation technology, transforming expert experience into logical rules understandable by DeepSeek.

IEPlat Tengyun has built-in over 100 diagnostic algorithm models from Zhongyun, is compatible with equipment and systems from multiple vendors, and supports integration with third-party systems. Through IEPlat, enterprises can securely display, share, analyze, and push equipment data on a unified platform, building a proprietary equipment database, and can view equipment monitoring and diagnostic results in real time via App or WeChat mini-program, instantly receiving alarm notifications, and comprehensively controlling the operating status of equipment.

• Intelligent diagnosis + manual diagnosis, providing 24/7 remote service

Zhongyun Technology has established a remote diagnostic center consisting of more than 20 vibration analysis experts, serving as a smart "external brain" for enterprise equipment management, providing comprehensive equipment maintenance services such as regular remote equipment checkups, real-time monitoring, diagnostic analysis, and on-site fault analysis.

The remote diagnostic center achieves 24-hour real-time status monitoring by connecting enterprise equipment, and combines Zhongyun's diagnostic algorithm models to perform equipment status assessment, intelligent early warning, and intelligent diagnosis. Once abnormal equipment data is detected, the center will promptly notify enterprise customers of the faulty equipment and possible causes through SMS, WeChat, public account push, etc. Analysts can also provide detailed diagnostic reports and fault analysis to help enterprises accurately understand the equipment status, achieve rapid fault location and repair, and accurately grasp the equipment health status. First-line inspection and maintenance personnel can also view the equipment status in real time through the App or WeChat mini-program. This not only reduces the enterprise's daily reliance on equipment inspection personnel, improves fault response speed, but also saves the cost of hiring additional analysts.

It is worth mentioning that, to help enterprises more efficiently collect various operating status data from on-site equipment and promptly process and diagnose the data, Zhongyun Technology has independently developed a series of intelligent sensor and edge intelligence terminal hardware products, including vibration, oil, current harmonics, and MEMS. The three-in-one sensor is widely favored. It can simultaneously measure five channels of signals: X, Y, and Z-axis vibration, sound, and temperature, achieving high-precision, fast, continuous, and synchronous acquisition of vibration in the axial, radial, and horizontal directions of the equipment, significantly enhancing the comprehensiveness and efficiency of data collection.

Deep cultivation of the industry, significant implementation and landing effects

The difficulty in implementing PHM projects lies not only in the complexity of the technology itself, but also in the fact that its implementation needs to be combined with the specific equipment and operating environment of the enterprise, making it difficult to simply replicate and presenting a "thousand enterprises, thousand faces" characteristic. However, Zhongyun Technology, with its deep technological accumulation, strong R&D capabilities, and rich implementation experience, has successfully served customers in more than 20 industries. It has collaborated with universities such as Harbin Institute of Technology, Huazhong University of Science and Technology, and Xi'an Jiaotong University, and jointly established a PHM laboratory, focusing on exploring and solving industry pain points and common problems.

Currently, Zhongyun Technology has formed solutions for various industries and fields, including cement building materials, petrochemicals, power energy, coal mining, iron and steel metallurgy, automobile manufacturing, water treatment, food processing, motors, pumps, and wind turbines, monitoring more than 50,000 pieces of equipment.

In the cement building materials industry, Zhongyun Technology has established in-depth cooperation with well-known cement building materials enterprises such as Huaxin Cement, HaiLuo Cement, China National Building Materials Group, China Resources Cement, and Jiumu Jidong. For example, a certain cement group, based on Zhongyun's complete set of hardware solutions and fault algorithm software, designed approximately 550 measuring points and manages approximately 105 pieces of equipment. Since its operation in December 2020, the group has issued early warnings for more than 100 mechanical failures, including loose motor bearings, rotor wear, and loose screws, preventing production accidents caused by equipment, achieving zero missed reports to date, reducing daily spot inspection workload by 60%, and reducing unplanned downtime by 30%.

In the petrochemical industry, there are a wide variety of equipment, and the equipment operating environment is harsh and dangerous. By monitoring the daily operating status of the equipment and pre-diagnosing faults, equipment failures can be effectively prevented, ensuring the effectiveness and pertinence of maintenance measures. For example, a certain refinery, through the equipment intelligent maintenance solution provided by Zhongyun Technology, has covered approximately 200 pieces of equipment with a total of 750 measuring points in two months of implementation, including important equipment such as vertical pumps, centrifugal fans, screw air compressors, reciprocating compressors, and turbine compressors. As of December 2023, the system has successfully issued more than 100 early warnings of equipment anomalies, timely discovering and handling potential faults, effectively avoiding unplanned downtime and production safety accidents caused by equipment failures.

In the automotive industry, the production equipment of automobile manufacturing enterprises is complex in structure and difficult to maintain, usually requiring professional technicians to track and monitor for a long time. Status monitoring can timely discover equipment problems; with the help of sensors and alarm rules, the burden of daily spot inspection can be effectively reduced, and work efficiency can be improved; combined with artificial intelligence algorithms, it can ensure the continuous and stable operation of equipment and reduce unplanned downtime. For example, a certain auto parts company, by introducing Zhongyun Technology's equipment predictive maintenance solution, has achieved a reduction in maintenance costs of approximately 30%-40%.

According to a study by the US National Science Foundation, by researching and deploying PHM technology, it can bring more than US$855 million in economic benefits to the industry, achieve near-zero unplanned downtime, and obtain more optimized maintenance strategies. As a pioneer in the domestic PHM market, Zhongyun Technology's full-range intelligent maintenance product matrix, especially its independently developed more than 100 high-precision diagnostic algorithm models, and its innovative service model integrating intelligent diagnosis and manual diagnosis, has greatly accelerated the application of PHM technology in enterprises, effectively reducing enterprise maintenance costs, reducing the incidence of failures, improving overall safety and economic benefits, and injecting strong impetus into enterprise transformation and upgrading.