With the implementation of modern production strategies, such as “Industry 4.0” and “Made in China 2025,” the development of high-end mechanical equipment is gradually reaching a high degree of parameterization, digitalization, networking, and intelligence [1]. High-end mechanical equipment no longer consists of traditional single devices, but is closely linked to the overall production process; that is, a variety of devices interact with each other and collectively form a complex system.

However, faults can lead to production delays throughout the system and even seriously threaten the safety of workers [2]. In addition, equipment operating in unmanned air or space vehicles (including space robots) cannot be repaired by humans if they break down [3]. Until recently, the typical method of addressing machine faults was primarily the “fault cure” method, which involves manually shutting down, inspecting, and repairing the equipment to restore normal operation. Accordingly, the maintenance cycle of the traditional “fault cure” method is time-consuming, and the quality of the maintenance depends on the skills of the technician [4].

The demanding performance requirements for high-end mechanical equipment have contributed to the development of artificial self-recovery theory and self-recovery regulation technology. Artificial self-recovery theory replaces the “fault cure” method of manually repairing mechanical equipment with the “fault self-recovery” method, in which mechanical equipment repairs itself. Specifically, machines are given the capability of eliminating or suppressing faults online through an intelligent decision and active control mechanism that is activated in the early stages of fault detection [5], [6], [7]. The self-recovery process does not require human participation, promotes the development of intelligent and unmanned high-end mechanical equipment, and provides effective technical support for achieving “Industry 4.0” and “Made in China 2025.”

The rest of this paper is organized as follows. In Section 2, the principles and history of artificial self-recovery theory are introduced, and the status of Chinese research on self-recovery technology is compared to that of international research in this area. Section 3 analyzes how artificial self-recovery empowers equipment to maintain its health autonomously, and summarizes self-recovery functions and equipment in detail. Section 4 describes an experiment in which the unbalanced vibration amplitude of a grinding wheel is reduced by 91.3% and the quality of the workpiece is improved by 40.0%, proving the value of artificial self-recovery theory and self-recovery regulation technology for engineering applications. Finally, Section 5 contains the conclusions of this study.

Artificial self-recovery theory was first proposed in 2003 by Jinji Gao [8] at the International Conference on Intelligent Maintenance. The theory originated from modern engineering practice and development requirements. Essentially, it is a scientific theory designed for dealing with the “unfaulted state,” which is a concept derived from a context in which modern mechanical equipment and Chinese medicine were juxtaposed. It integrates state monitoring, diagnostics, artificial intelligence, active and adaptive control, intelligent structure, and embedded technology. The purpose of the theory is to grant the equipment autonomous control, and self-recovery regulation technology is aimed at preventing and eliminating faults to enable self-recovery.

Jinji Gao was not only the first person to propose artificial self-recovery theory, but he was also the first person to carry out research on self-recovery regulation technology. In 2006 and 2011, he led a team to undertake two key projects of the National Natural Science Foundation of China. The team applied self-recovery regulation technology to compressor shaft displacement faults, turbine unit instability faults, the automatic balancing of high-end machine tools, and the control of unbalanced vibrations in aero-engines [9], [10], [11]. In one experiment, a self-recovery regulation system for unbalanced vibrations was applied to a cylindrical grinding machine. Applying the theory to the system effectively suppressed the vibrations of the grinding machine to the submicron level without human intervention. Simultaneously, the surface quality of the workpieces processed by the grinding machine was improved by 25%-40%. This indicates that artificial self-recovery theory and self-recovery regulation technology can ensure the efficient and stable operation of mechanical equipment in various industries.

As a result of these experiments, in 2011 the Chinese Academy of Engineering suggested that the medium- and long-term development strategy of Chinese engineering technology should focus on ensuring that large specialized equipment incorporates fault prediction and self-recovery regulation technology [12]. Ministry of Industry and Information Technology of the People’s Republic of China also pointed out in the “Industrial Internet and Safety Production” Action Plan (2021-2023) that it is necessary to apply intelligent control more comprehensively and to accelerate the promotion of industrial solutions such as self-recovery regulation technology [13]. This demonstrates that the concept of artificial self-recovery has been affirmed by the Chinese engineering community, which has created promising prospects for the high-end mechanical equipment industry.

While Jinji Gao was the first to successfully apply artificial self-recovery theory to high-end machine tools and compressors, other scholars have also carried out research on artificial self-recovery theory and technology in different fields. For example, Xuedong Chen groups [14], [15] from the China National Machinery Industry Corporation has examined the feasibility of applying artificial self-recovery theory to the maintenance of pressure vessels. In fact, a self-recovery maintenance method for pressure vessels based on actual maintenance cases has been proposed in accordance with pressure vessel design and manufacturing methods, and it has been combined with digitization and networking techniques. In addition, Hong Liu group [16] from the Harbin Institute of Technology has investigated the current status of the application of artificial self-recovery theory to the self-repairing capabilities of spacecraft and space robots, and has analyzed the challenges associated with the specificity of their environment. Furthermore, He Lin group [17], [18] from the Naval University of Engineering has evaluated the problem of reducing the hull vibrations of large ships. They proposed an intelligent vibration-isolating device based on structural self-adaptation and ship hull control methods, and combined it with artificial self-recovery theory. Moreover, Xinping Yan group [19], [20] from the Wuhan University of Technology has developed a mechanistic model of the bearing state and has tracked the progress of self-recovery regulation research. Similarly, Zengliang Gao group [21], [22] from Zhejiang University of Technology has reviewed case studies of pressure vessels and the progress of research on advanced structural design and self-recovery theory.

From 2021 to 2022, under the guidance of the Chinese Academy of Engineering, the China Association of Plant Engineering, and other relevant institutes, the Beijing University of Chemical Technology hosted the first and second academic forums on Artificial Self-Recovery and Autonomous Health of Machine. Nine academicians and more than 20 domestic experts were invited to contribute reports, and the number of online participants reached more than 40  000. The current status of self-recovery regulation technology and high-end mechanical equipment in different fields was presented, which significantly enriched the subsequent development of artificial self-recovery theory. Typical applications of artificial self-recovery theory are illustrated in Fig. 1.

The United States and other manufacturing nations are also conducting self-healing research. For example, in 2007 the National Aeronautics and Space Administration (NASA) began improving the automatic adjustment capability and adaptive speed of control systems through the Integrated Resilient Aircraft Control project [23]. In addition, in 2012 the National Science Foundation began developing self-healing theories and analyzing principles of self-healing systems through the Failure-Resistant Systems project [24]. Furthermore, in 2013 researchers at the University of Illinois developed a rubber adhesive that can cure at high temperatures and enhance the toughness of the material, which could increase the autonomous repair capabilities of spacecraft structures [25]. In 2015, Alessandro Casavola group [26] enhanced the automatic fault regulation capabilities of aerospace control systems by combining fault diagnosis and fault-tolerant control.

The literature analysis provided above indicates that the self-healing theory proposed by American researchers is designed only for the structural or functional maintenance of the flight control systems in aircraft, which has a narrow range of practical applications [27], [28]. Artificial self-recovery theory has a wider range of engineering applications, which are relevant for spacecraft, pressure vessels, high-end machine tools, and other technologies. Artificial self-recovery theory focuses more on active and automatic regulation, such as through a robotic arm or self-recovery regulation actuators, to find the optimal treatment strategy for rectifying a fault. Adaptive regulation has the disadvantages of being uncontrollable and irreversible, but active regulation allows the regulation process to be more controllable and effective. Mechanical equipment systems achieve autonomous health by eliminating or suppressing faults online via self-recovery regulation.

The human body is made up of various organs and systems; similarly, machines are devices made up of various parts and components. Accordingly, there are commonalities between the bodies of humans and the machines they create. Likewise, there are also many similarities between human diseases and machine faults. When diagnosing human diseases, it is necessary to identify the unhealthy organs or tissues and prescribe the appropriate medication. Similarly, when preventing machine faults, it is necessary to account for the working conditions and status of machines through real-time monitoring. Thus, fault prevention in machines is analogous to the diagnosis of diseases in humans; therefore, the latter can inform artificial self-recovery theory.

The World Health Organization has proposed that the focus of medical research in the 21st century should shift from disease treatment to disease prevention and the assurance of human health. The ultimate purpose of medical diagnosis is to anticipate discomfort, prescribe the correct medication, and avoid physically damaging patients. Therefore, machine fault prevention research should also focus on predicting faults, timely self-restoration and adjustment, and providing solutions. These three major steps are the core idea behind artificial self-healing theory and self-recovery regulation technology. When an abnormal working condition occurs, the self-recovery system handles the fault by quickly analyzing its cause, identifying the faulty component, and formulating treatment measures. By drawing on medical science concepts, self-recovery regulation technology can play a very productive role in the development of modern high-end mechanical equipment. As shown in Fig. 2, the purpose of machine fault diagnosis is to gradually transition from the need for manual troubleshooting to the construction of self-predicting and self-repairing mechanical equipment.

The field of artificial self-recovery theory is broad, consisting of self-recovery technology, substitution technology, and self-protection technology. As a new direction in bionics, self-recovery regulation technology relies on the analysis of fault mechanisms and risks, and gives machines the ability to act spontaneously and maintain an optimal state through bionic design. The technology allows machines to detect abnormal conditions and repair them while operating based on existing methods for diagnosing and repairing machine faults. Artificial self-recovery theory discards the traditional method of stopping the machine for manual troubleshooting and incorporates a self-recovery function into the inherent operating procedures of the machine, ensuring autonomous health. A number of faults and their corresponding self-recovery functions, powers, and mechanisms are shown in Table 1 [10], [29], [30], [31], [32], [33], [34], [35].

In recent years, the authors have been devoted to researching self-recovery regulation technology for high-end mechanical equipment, and have developed self-recovery regulation capabilities for grinding wheel vibrations, compressor shaft displacement faults, and turbine unit destabilization faults. In this study, the authors tested the practical engineering applications of self-recovery regulation technology by experimenting on an unbalanced vibration fault in a high-end grinding wheel, using a self-recovery regulation system to repair the fault and machine the workpiece [10].

The operation principle of the self-recovery regulation system is shown in Fig. 3(a) [36]. The vibration acceleration and speed sensors measured the vibration and speed signals of the grinding wheel, respectively, and other corresponding signals in real-time. These signals were transmitted to the control device. The digital acquisition module then processed the real-time data and identified the unbalanced vibration parameters of the grinding wheel. The control device quickly located the initial imbalance of the grinding wheel based on the unbalanced vibration targeting control method, and issued a command to instruct the self-recovery actuator. In response, the actuator gradually changed its mass distribution to produce a vector that could offset the initial imbalance, stabilizing the grinding machine and enabling the rotor vibration amplitude to satisfy the operating requirements. Finally, to complete the process, the compensation vector and initial imbalance vector of the rotor were completely offset, causing the vibration to subside. After applying this system to an actual high-end grinding machine, which is shown in Fig. 3(b), the vibration amplitude of the grinding machine was reduced by 91.3% and the quality of the machined workpiece was improved by 40.0%. These results verify the effectiveness of the self-recovery regulation system and prove the feasibility of artificial self-recovery theory for a wide range of engineering applications.

According to the latest literature reports, other universities in China are still in the theoretical research or laboratory test stages, and no practical application of self-recovery regulation technology has been reported [37], [38].

Artificial self-recovery theory is an important guide for promoting intelligent and self-recovering mechanical equipment. It has been applied to problems such as the control of unbalanced vibrations in high-end machine tools and aero-engines, and has proven to be effective. In this study, artificial self-recovery theory was used to reduce the unbalanced vibration amplitude of a grinding wheel by 91.3% and improve the quality of the workpiece by 40.0%, proving the value of self-recovery regulation technology for engineering applications.

However, current self-recovery regulation technology primarily relies on external active drives that assist in eliminating faults, which is equivalent to adding auxiliary components to the original structure. Additionally, current self-recovery regulation technology needs to be combined with existing equipment, but this requires installing the technology in a position that accommodates it, which limits the impact of the self-recovery mechanism.

The future development of artificial self-recovery theory should focus on incorporating self-recovery regulation systems into the early stages of equipment design and manufacturing. Equipment with self-recovery regulation systems can solve most faults before any failure occurs. However, self-recovery regulation technology installed in complex mechanical systems requires not only parametric adaptation to changes, but also structural adaptation to faults. This new and transformative technology has a broader scope than traditional control systems owing to the increased complexity of the problems, which introduces significant challenges for developing engineering solutions. Nevertheless, its promotion and application will significantly improve productivity, deliver substantial economic benefits to various fields, and bolster a new technological revolution.

This work was supported by Natural Science Foundation of Beijing (3212010) and National Natural Science Foundation of China (51875031 and 52242507).