Review of On-Orbit Assembly Technology with Space Robots

Abstract

With the accelerated pace of human space exploration and the progress of other related researches, there is an increasingly urgent demand for space infrastructure, equipment, and diversified spacecraft construction for space missions, and how to efficiently, intelligently, and autonomously build corresponding facilities and equipment on orbit according to the functional requirements of different missions has become a great challenge in the field of space technology research. As an important means of automated manufacturing, the construction of on-orbit assembly systems centered on space robotics has become an emerging development trend. In view of its importance, space agencies and research institutes have successively proposed and developed a series of related programs. In order to comprehensively understand the progress of on-orbit assembly with space robots (OASR) and scientific problems involved, this paper investigates the current status of research and technological development in OASR. Firstly, the significance of OASR for space exploration and other space missions is analyzed. Secondly, the existing classification forms of on-orbit assembly are outlined and a classification idea is proposed from the point of view of the combination of space robot motion capability and assembly goals. Thirdly, the research and development status of OASR in the United States, Europe, Canada, Japan, and China is investigated. Then, based on a review of the literature on space robots to realize on-orbit assembly in space facilities, some of the key technologies involved are reviewed and discussed. Finally, this paper discusses and looks ahead to the future development trend and application prospect of the technology of OASR, reveals and explains the crucial position it occupies as well as the important role it can play in the process of human space exploration, and is expected to provide useful references for the in-depth research and development of future on-orbit assembly technology.

1. Introduction

The technology of on-orbit assembly with space robots (OASR) serves as the primary means for on-orbit service and maintenance, as well as for the construction and deployment of space equipments or infrastructure. In view of the broad spectrum of space robots, the present focus is on the investigation and elaboration of space robot forms equipped with robotic arms as the primary operational actuators. Under this circumstance, in order to facilitate on-orbit maintenance, module replacement, fuel refueling, and other forms of on-orbit services targeting satellites, countries have engaged in carrying out research and verification of relevant fundamental key space technologies [1]. Notable examples include the “Orbital Express” and “Phoenix” of the United States, the “ETS-VII (Engineering Test Satellite-VII)” of Japan, and the “ROTEX (Robot Technology Experiment)” and “DEOS (Deutsche Orbital Servicing Mission)” of Germany. These projects have all adopted space robotics as the core technology program for the on-orbit maintenance and module replacement of satellites [2]. In terms of on-orbit fuel refueling, NASA employed the Canadian Space Agency’s Dextre robot to conduct on-orbit technology validation of fuel refueling [3]. Similar validation projects include the MEV-1 (Mission Extension Vehicle-1) project developed by Northrop Grumman [4], the Tianyuan-1 project created by the National University of Defense Technology of China [5], and the on-orbit fuel replenishment verification of Tiangong-2 by Tianzhou-1 [6]. In the context of on-orbit assembly, typical examples are the assembly of the Hubble Space Telescope [7] and the construction of large space antennas [8]. In regard to space infrastructure, as shown in Figure 1, the advancement of deep space exploration has resulted in an increasingly urgent necessity for the on-orbit construction of large-scale space platforms and infrastructure. This includes, but is not limited to, space solar power stations [9], on-orbit fuel supply stations [10], deep space exploration transit stations [11], space vehicle home ports, and extraterrestrial residence bases [12,13].

As a space-enabling technology, the technology of OASR plays an instrumental role in the promotion of the development of space science. And it has constituted a significant element of the space sector since the 1960s. In recent years, in conjunction with the rapid advancement of space exploration, countries have also enacted policy measures to accelerate the development of associated technologies. The National Space Strategy of the United States provides guidance on a number of key areas, including space operations, basic space capabilities, and the commercialization of space. Russian Federation Space Program 2016–2025 explicitly calls for the advancement of research in the domain of sophisticated technologies, encompassing the realms of novel space machinery fabrication, electronic equipment production, and material science processes. The European Space Strategy of the EU places an emphasis on the integration of European space. The Fourth Medium- and Long-term Development Plan of Japan emphasizes on the expansion of research activities within the domains of space transportation, space system maintenance, and space science and exploration. The white paper of China’s Spaceflight 2021 highlights the advancement of space technologies and systems, which are being utilized to establish the groundwork for space infrastructure and deep space exploration. It is anticipated that the pace of technological development in the space sector will increase in the future. The development and breakthrough of technology utilising space robots for on-orbit assembly is expected to revolutionise the existing paradigm of space construction and maintenance, expand the boundaries of human exploration, and have a significant impact on future development.

The mode of OASR presents a number of advantages over the conventional approach of ground assembly followed by launch vehicle launch [14]. Firstly, the conventional method for launching facilities and equipment from the ground into space, which employs launch vehicles, is constrained by a number of factors, including the capacity of the rocket and the volume of the payload envelope. Moreover, in order to withstand the overloads of the launch phase, the payload structures are overstrengthened, and these reinforcements are useless for their operation in the space environment. Compared to on-orbit assembly forms, this reduces the cost-effectiveness of traditional deployment methods. The method of OASR is a technique for the construction and integration of space modules, ranging from small functional structures to large-scale infrastructures in space. This mode permits the dimensions of the target to be independent of the launch vehicle’s spatial envelope, reduces the structural weight necessary to withstand the mechanical environment of the launch phase, and significantly increases the material utilization, thus enabling the assembly and manufacture of very large space structures [15]. Secondly, the flexible and elastic construction of space equipment and facilities can be achieved through the addition, replacement, and reconfiguration of functionally definable modules. This allows for the realization of on-demand and emergency assembly and manufacturing of functional equipment, complemented by timely on-orbit maintenance, which can further extend the lifespan of the target equipment and, at the same time, achieve the continuous space presence of specific functional equipment or facilities. Moreover, this mode has the potential to transform the assembly and manufacturing process, enabling the exploration, assembly, and manufacture of structures that are not feasible on Earth due to the influence of gravity. Additionally, in consideration of the challenging conditions in space, it can be tentatively concluded that the deployment of space robotic systems in space assembly and manufacturing tasks will have the dual advantage of enhanced precision and efficiency, while also mitigating the risks associated with manual labor, as compared to manual space assembly. As previously stated, the utilization of space robots for on-orbit assembly not only expands the spatial scope of the original aerospace structural design, enriches the manufacturing process and means, and extends the boundaries of human capabilities, but also provides more possibilities for the construction of future space megastructures, thus establishing a technological foundation for the exploration of deep space by humankind.

In the aforementioned context, this paper aims to provide an overview of the current state of development and related technologies for OASR. This paper is organized as follows: Section 1 presents an overview of the research background and significance of OASR. Section 2 delineates the proposed conceptualization of classification methodologies within the context of the diverse classification strategies that have been devised for OASR. Section 3 is dedicated to an examination of the prevalent space robotic arm systems currently employed in space missions around the globe. Section 4 is concerned with an investigation of the current status of research and development in the field of OASR. Section 5 presents a comprehensive analysis of the primary research institutions and research hotspots in the context of a literature review. Section 6 provides a succinct overview of the fundamental technologies pertinent to OASR, while also elucidating the challenges that have emerged based on the aforementioned status. Section 7 delves into the technological trends associated with OASR, and a comprehensive summary and outlook are presented in Section 8.

2. Classification of On-Orbit Assembly

The technology of OASR represents the fundamental technology for future space exploration and constitutes an essential means of realizing the rapid and cost-effective construction of space facilities and equipment. The relevant research can be traced back to the 1990s, and scholars have proposed various classification methods from different dimensions according to the different forms of on-orbit assembly, assembly realization methods, complexity of the objects, and spatial location of assembly execution. However, the existing classification methods only start from the single dimension of assembly objects or actuators, which has certain limitations and is difficult to comprehensively cover the diversity and complexity of on-orbit assembly technologies. Therefore, this paper proposes a combined classification method, which is expected to reflect the characteristics of OASR more comprehensively and meticulously, and provide a reference for the in-depth study of the technological needs and directions under different assembly scenarios. In this section, we will first introduce the relevant classification forms and then present our thoughts on the classification with the focus of this paper.

2.1. Target and Task Characteristics-Based Classification

According to the scale and complexity of on-orbit assembly targets, some scholars suggest that forms can be classified into five tiers: spacecraft combination, function expansion, whole star assembly, module assembly, and on-orbit manufacturing [14]. Spacecraft assembly encompasses the processes of docking and on-orbit integration of functional and relatively independent spacecraft. The International Space Station (ISS) is an exemplar of this hierarchy, having undergone several module launches and assemblies since 1998, in addition to the incorporation of ancillary equipment such as trusses and space robot arms, before attaining its current configuration. A case in point of function expansion is the “Orbital Express” project, which was proposed by DARPA. This project verifies the docking of targets and the on-orbit replacement and assembly of components by replacing parts for failed satellites. The GOAT project, proposed by the Surrey Space Center of the UK, represents the typical example of a whole star assembly. It realizes modular assembly by enabling the assembly unit with reflector to complete the rendezvous and docking with the main structure. A typical project of module assembly is the CIRAS program proposed by NASA Langley Research Center, which consists of a precision manipulator robot, IPJR, and a TALISMAN arm with an arm span of 15 m, and they have been verified in the assembly of truss structures and solar arrays. The archetypal exemplar of on-orbit manufacturing is the SpiderFab program, which was proposed by NASA. The program’s fundamental component is a multi-arm robot that is capable of 3D printing in space. This enables the in situ manufacturing of structural units through 3D printing, which are subsequently assembled into a larger space structure.

In light of the intricate nature of spacecraft assembly tasks, scholars at the Shenyang Institute of Automation of the Chinese Academy of Sciences have put forth a four-level division method for spacecraft on-orbit assembly [16]: (1) element-level assembly, where two or more independent spacecraft are assembled in space with a minimum of complexity into a larger space structure; (2) module-level assembly, which involves the docking and assembly of a capsule or module to create an independent spacecraft or to extend the reconfiguration of its functionality; (3) component-level assembly, which is a sequential process involving the assembly and integration of components, which often requires a significant number of connection-related operations and interactions; and (4) part-level assembly, which is a multi-stage process that begins with the initial parts and culminates in the assembly of components. This method is analogous to the assembly of a computer, which involves the integration of various components, including power supplies, heat sinks, and other parts.

From the perspective of the dimensions of the assembly object, the first division is more discernible. In the absence of consideration of the level of on-orbit manufacturing, the meanings of the above divisions are also found to be highly analogous. However, when considering the complexity and functionality of the object, it becomes evident that the most complex components are concentrated in the third and fourth levels. Moreover, the definition of a functionally independent module itself is characterized by a certain degree of nested attributes. As a result, it is more challenging to ascertain the degree of complexity, whether it be the entirety of the star assembly and module assembly, as in the first method described, or the part-level assembly and component-level assembly, as outlined in the second approach. Therefore, the criteria for differentiation require clarification.

2.2. Assembly Realization-Based Classification Method

On-orbit assembly can be categorized into manned and unmanned methods, depending on how it is implemented. Manned on-orbit assembly refers to the form with the help of tooling assistance dominated by human beings (mainly cosmonauts), as shown in Figure 2. The replacement of the heat shield and deployment of solar panels on the Skylab space station of the United States [17], the assembly of the truss structure and experimental module on the MIR space station of the Soviet Union, the in-orbit assembly and maintenance of the Hubble telescope and the replacement of the sensors, and some of the on-orbit assembly projects on the ISS [18] were all carried out through the leadership of astronauts. The manned on-orbit assembly method requires that astronauts engage in extravehicular activities. However, the implementation of manual on-orbit assembly is more challenging due to the size of the assembled target, the extravehicular environment, and the physical fitness of the astronauts. To illustrate, the Langley Research Center (LaRC) of the United States has conducted a series of research and verification projects on technologies related to manual assembly of large-scale space structures in ground gravity, ground simulation of zero-gravity, and space zero-gravity environments. The findings of these studies illustrate that (1) the use of mobile workstations and pendulum robotic arms to facilitate the assembly process can mitigate the effects of astronaut fatigue, (2) meticulous mission planning can minimize the time spent by astronauts in idle states during assembly, thereby enhancing operational efficiency, and (3) the deployment of construction methodologies that combine manual and robotic-assisted techniques can reduce costs and risks while simultaneously achieving functional diversity in spacecraft structures. Although the form of manned on-orbit assembly is well-suited to assembly tasks with small task volumes, short time frames, and relatively simple environments, it is challenging to meet the requirements of high-precision, long-term, and high-efficiency assembly of large and complex structures in space.

Unmanned on-orbit assembly refers to the construction of space facilities or equipment that is realized by technical means, such as space robotics. Various organizations, including NASA, ESA, and JAXA, have conducted research on related technologies. The primary methods employed include the use of remote operation and autonomy technology. For example, Germany’s Space Factory 4.0 employs low-latency remote operation technology to achieve precise space assembly tasks, while Boeing has conducted verification of space robot self-assembly technology. In comparison, the technology of utilizing a space robot arm for module operation was initiated at an early stage and has reached a high level of technical maturity. This is exemplified by the “Orbital Express”, which employed a space robot to complete the power supply of rendezvous satellites and the replacement of the attitude control module. Carnegie Mellon University designed the Skyworker space structure, which is attached to a mobile robotic system, for use in the assembly of space trusses. NASA Jet Propulsion Laboratory designed a six-legged walking robot, LEMUR, for use in narrow space complex fine assembly and overhaul tasks. Japan conducted verification of modules replacement through the robotic arm in 1997 on the engineering experimental satellite ETS-VII.

The current categorization of manned versus unmanned or manual versus automated has a wide range of applicability. In particular, the domain of unmanned assembly is witnessing the emergence of novel conceptual solutions on a regular basis. This represents a highly active and pivotal area of current research, as well as the central focus of the present study.

2.3. Combination-Based Classification Method

In contrast to the aforementioned classification form, this section offers a reflection on the assembly classification form from the vantage point of the relative movement relation between the space robot and the task environment, as shown in Figure 3. In contrast with the classification based on complexity, the current method does not result in the formation of formal nesting due to the issue of assembly object size, reducing the impact of assembly complexity on the classification process to a certain degree. In addition, compared with the assembly realization-based classification strategy, the current method is no longer limited to the differences between assembly subjects, but focuses on the dynamic relationship between robots and the task environment, providing a more comprehensive coverage of the various factors involved in the assembly process.

In this section, OASR are classified into three categories: fixed space assembly, semi-fixed space assembly, and floating space assembly. Fixed space assembly forms are defined as those in which the task space of robot remains relatively fixed to the robot’s mounting base. An example of this would be a robotic arm fixedly mounted within the space station’s hull to perform in-cabin assembly tasks. In contrast, semi-fixed forms are those in which the robot’s mounting base and task space can be moved along a lesser number of degrees of freedom than the number of dimensions of task space. This can be exemplified by assembly robotic systems that move along a space truss. In floating space assembly forms, the robot is mounted on a floating base that can be freely moved within the mission space. This type of assembly is exemplified by satellite robots or freely movable space assembly modules. Furthermore, the assembly object can be classified as either an array assembly object or a non-array form based on its composition. An array assembly object exhibits a relative unity of constituent components and a regular or repeatable installation mode. In contrast, a non-array object is an assembly module that lacks the structural similarity of the target object.

In accordance with this classification methodology, a general on-orbit assembly task can be conceptualized as a conjunction of space robots linked by means of space assembly interaction technology and the object to be assembled. As illustrated in Figure 3, a variety of combinations are applicable to diverse assembly tasks. Fixed assembly robots, for instance, are capable of performing the assembly of intricate and precise small components, expendable components, standard components, and structural connectors. Semi-fixed assembly robots are capable of completing the structural assembly of the main body of large space facilities for non-arrayed structures or of completing the structural assembly of standardized components for arrayed objects. Floating robots, meanwhile, are able to realize the mobile docking and installation of functionalized structures.

3. Overview of Space Robotic Arms

As an important form of space robotics, the space robotic arm usually has more degrees of freedom, its movement is flexible and it can be equipped with diversified functional end effectors, which can play an important role in space assembly, on-orbit service, and extra-vehicular assistance. For this reason, the United States, Canada, the European Union, Japan, and China have invested huge human and material resources to promote the research, development, and application of space robotic arm technology. This section introduces the major space robotic systems with a focus on space robotic arms.

3.1. Status of North American Space Robotic Arm Research

The development of space robotic arm technology in North America is dominated by the United States and Canada. The United States pioneered the concept of a space robotic arm in the 1970s, and Canada’s SPAR Corporation then made it a reality [19].

In 1981, Canada’s SPAR developed the Shuttle Remote Manipulator System (SRMS, known as Canadarm1), shown in Figure 4a, which plays a key role in payload deployment and recovery, satellite maintenance and servicing, extravehicular activity guidance and assistance, and ISS construction and assembly [20,21]. Since the SRMS has a limited operating range and cannot cover all the target locations of the ISS, Canada further developed the Space Station Remote Manipulator System (SSRMS, known as Canadarm2) based on the SRMS, as shown in Figure 4b, which was deployed into space in 2001. With four times the payload capacity of the SRMS, the SSRMS can not only easily reach the ISS mission site, but can also combine large load handling with the capture function of visiting vehicles [1].

In 2007, Canada developed a more agile special purpose dextrous manipulator (SPDM, known as Dextre) robotic arm, which was launched to the ISS in 2008. SPDM is a dual-arm robotic system, as shown in Figure 4c, with each arm having seven degrees of freedom and equipped with an orbital replacement unit (ORU) tool change mechanism that can be fine-tuned through real-time force feedback and other features [22]. SSRMS, SPDM, the Mobile Remote Servicer Base System (MBS), and the Mobile Transporter (MT) together form the ISS Mobile Servicing System (MSS), as shown in Figure 4d, which is by far the most complex robotic system on the ISS. It can move anywhere along the truss orbit outside the ISS with the MBS and MT, and with the high payload capacity of the SSRMS and the highly flexible and fine-tuned operational capability of the SPDM, it can provide service support for the assembly and maintenance of the ISS [23].

Canada is currently working with NASA on the next generation space robotic arm, Canadarm3, as shown in Figure 4e. Canadarm3 is expected to be operational in 2028 for the autonomous capture, maintenance, and repair of lunar rovers in the Lunar Gateway program [24,25]. In addition to the above space robotic arms, the United States has developed robotic astronauts such as Robonaut2, shown in Figure 4f, which has been technologically validated aboard the ISS [26].

3.2. Status of European Space Robotic Arm Research

Research and development of space robotic arm technology in Europe is concentrated in two institutions: the European Space Agency (ESA) and the German Aerospace Center (DLR, Deutsches Zentrum für Luft-und Raumfahrt).

The European Robotic Arm (ERA), shown in Figure 5a, was designed as a symmetrical structure to realize its “walking” function on the ISS, which can be used for extravehicular inspections of the station and payload transfer operations [27,28]. In 2021, ERA complemented the installation of solar arrays on the Russian segment of the ISS. Another space robot under development at ESA is the Eurobot assisted three-armed robot, shown in Figure 5b. The Eurobot consists of three seven-degree-of-freedom DexArm arms for greater mobility and functionality. Currently, Eurobot has completed underwater testing and will be used for extravehicular operations on the ISS in the future [29].

DLR has developed a lightweight space robot, ROKVISS, shown in Figure 5c, which can perform teleoperation experiments over long distances and with large time delays [30,31,32]. DLR is currently developing the Compliant Assistance and Exploration SpAce Robot CAESAR, shown in Figure 5d. CAESAR uses lightweight design techniques to capture and stabilize uncooperative satellites for assembly, maintenance, repair, and debris removal [33].

In addition, Russia has developed a humanoid robot, the Skybot F-850, which can be used to grasp electrical drilling tools, install electrical connectors, and verify human–robot interaction [34].

3.3. Status of Japanese Space Robotic Arm Research

The development of space robotic arm technology began early in Japan. Since 1975, a series of engineering test satellites have been successively launched, among which the ETS-VII satellite launched in 1997 was equipped with a space robotic arm system, which is also the world’s first satellite with a space robotic arm system.

In 2009, Japan installed the JEMRMS (Japanese Experiment Module Remote Manipulator System), a remotely operated robotic arm, in the Japanese Experiment Module segment of the ISS, as shown in Figure 6a. JEMRMS consists of a main arm (MA), a small flexible arm (SFA) and a control station, and is used primarily for servicing loads in the Japanese segment. The MA is 10 m long and has six degrees of freedom, which can easily grasp the object to be operated and also has the function of compliant operation, while the SFA can cooperate with the main arm to perform the fine operation tasks [35,36].

In 2021, the GITAI S1 robotic arm, developed by the Japanese space robotics company GITAI and shown in Figure 6b, performed an in-cabin assembly demonstration mission on the ISS to verify the arm’s ability to autonomously perform fine operations such as switching operations and unplugging and plugging interfaces [37]. The company has now developed an autonomous dual robotic arm system, S2, shown in Figure 6c, which has completed verification tasks such as ORU maneuvering, flexible material manipulation, and fastener attachment/detachment outside the ISS in March 2024, with a technology maturity level of 7, and is scheduled to achieve on-orbit service in 2026 [38].

3.4. Status of Chinese Space Robotic Arm Research

China’s exploration of space robot arm technology began in the 1990s, after a long period of technical research, has constructed a relatively complete space robot technology system. The Chinese Space Station Remote Manipulator System (CSSRMS) consists of two parts, as shown in Figure 7a: The core module manipulator (CMM), with a maximum working radius of 10 m and a load of up to 25,000 kg, also has a “crawling function” and can perform important tasks such as transferring between modules, assisting astronauts’ activities in and out of the module, handling extravehicular cargo, extravehicular status checking, and extravehicular maintenance of large-scale equipment. The experimental module manipulator (EMM), with a maximum working radius of 5 m and a payload of up to 3000 kg, is primarily responsible for servicing the experimental payload, supporting astronaut extravehicular activities, and performing extravehicular status inspection. CSSRMS has played an important role in the independent and autonomous construction of the Chinese space station [39].

In addition, in the early stage of technology verification of space robotic arm, the Harbin Institute of Technology developed the Test-7 space robotic arm based on modularized design concepts, as shown in Figure 7b, and completed the scientific test of space maintenance technology for on-orbit capture and target operation in 2013 [40]. In 2016, the China Academy of Launch Vehicle Technology (CAVT) developed the Oceanus-1 space robotic arm, as shown in Figure 7c, which further validated technologies such as active space debris removal and non-cooperative target detection and capture [41]. The 6dof lightweight space robotic arm developed by the Harbin Institute of Technology in the experimental module of Tiangong-2, as shown in Figure 7d, is equipped with a five-finger humanoid dexterous hand at the end, focusing on the verification of human–machine synergy in orbit, remote operation, dexterous grasping, maintenance in orbit, and other related technologies [42]. In the direction of the miniaturization, modularization, and multifunctionality of on-orbit assembly, the Shenyang Institute of Automation also developed a deployable robotic arm Cubot, which occupies only 1U (10 cm × 10 cm × 10 cm) of space, as shown in Figure 7e, and is capable of performing tasks such as space debris cleaning and on-orbit assembly and maintenance [43].

3.5. Summary of Space Robotic Arm Parameters

As an important technical means for the implementation of on-orbit assembly, the mechanical characteristics and parameter indices of the space robotic arm have a great influence on the on-orbit assembly tasks performed. In this regard, this part summarizes the main parameters of space robot arms developed by major domestic and foreign space agencies and institutes in the form of a table, as shown in

Table 1. Summary of the main parameters of the space robot arm.

No.	Name	Country	Year	Dof	Length/m	Maximum Load/kg	Accuracy	Current State
1	SRMS	Canada	1981	6	15.2	26,600	50.8 mm, 0.1°	On orbit
2	SSRMS	Canada	2001	7	17.6	116,000	45 mm, 0.71°	On orbit
3	SPDM	Canada	2007	7 + 1 + 7	3.5	600	6 mm	On orbit
4	Canadarm3	Canada	2012	7	8.5	——	——	Research
5	Robonaut2	America	2011	——	0.8	9	——	Returned
6	ROKVISS	German	2005	2	0.5	——	——	Ended
7	CAESAR	German	2018	7	3.1	——	——	Research
8	ERA	EU	2021	7	11.3	8000	5 mm, 1°	On orbit
9	Eurobot	EU	2003	3 × 7	——	——	——	Research
10	Skybot F-850	Russia	2019	——	——	——	——	On orbit
11	ETS-VII	Japan	1997	6	2	——	——	Ended
12	JEMRMS	Japan	2009	MA: 6	MA: 10	MA: 7000	MA: 50 mm, 1°	On orbit
				SFA: 6	SFA: 2	SFA: 300	SFA: 10 mm, 1°
13	GITAI S1	Japan	2021	8	1	——	——	On orbit
13	GITAI S2	Japan	2023	7	1.5	——	——	On orbit
14	CSSRMS	China	2021	CMM: 7	CMM: 10.2	CMM: 25,000	CMM: 45 mm, 1°	On orbit
				EMM: 7	EMM: 5	EMM: 3000	EMM: 10 mm, 1°
15	Test-7	China	2013	6	——	——	——	On orbit
16	Oceanus-1	China	2016	6	——	——	——	On orbit
17	TianGong-2	China	2016	6	——	——	——	On orbit
18	Cubot	China	2023	7	0.3035	——	——	Research

.

4. Status of On-Orbit Assembly with Space Robots

As the pace of space exploration accelerates, the structures of space facilities, equipment, spacecraft, etc., urgently needed for space missions are becoming more complex, while their performance and technological level are constantly improving. In this context, the development of an on-orbit assembly and manufacturing system with space robots at its core will not only provide strong logistical support for existing space facilities but also provide technical support for further human exploration into deep space in the future. Based on this, this section focuses on the issue of on-orbit assembly with space robots, taking the three types of space assembly forms proposed in this paper as the starting point, and explores, combs, and elaborates the relevant on-orbit assembly projects around the world.

4.1. Status of Research on Floating Space Assembly

On-orbit assembly with floating space robots has the advantages of unlimited working range, strong adaptability to complex tasks, and in situ assembly and manufacturing in clusters, etc. All major space agencies in the world have conducted exploration, experiments, and application projects to verify the assembly technology with floating space robots.

4.1.1. Status of Floating Space Assembly in North America

Since the 1990s, the U.S. and Canada have been conducting research on the technology of on-orbit assembly with space robotics as the core of relevant projects, in which the representative research institutions are NASA, DARPA, GSFC (Goddard Space Flight Center), as well as MDA, Boeing, and other companies.

In 2003, Boeing proposed the Autonomously Assembled Space Telescope (AAST), an autonomous on-orbit assembly project for a 10 m-class optical space telescope, as shown in Figure 8. The project focuses on space robotic pre-alignment and autonomous assembly of the basic components and main mirrors of space optical telescopes [44,45], and also explores the issues associated with payload and main spacecraft assembly and mirror aperture fabrication.

In the same year, the first satellite system of the Experimental Satellite Series (XSS) project, jointly conducted by the U.S. Air Force Research Laboratory (AFRL), the Space and Missile System Center (SMC), and the Naval Research Laboratory (NRL), was launched, as shown in Figure 9a. The goal of the project is to develop a fully autonomous microsatellite robotic system capable of on-orbit inspection, rendezvous and docking, and close maneuvering around orbiting objects. Among them, XSS-10, developed and completed by Boeing, focused on verifying technologies such as spacecraft-to-spacecraft proximity control and autonomous navigation [46,47]. The XSS-11, developed by Lockheed Martin and shown in Figure 9b, was launched in April 2005 and combined autonomous rendezvous control and in-space docking technology to explore the feasibility of low-cost space integration and manufacturing of satellite technology [48]. The future XSS-12 will be used to validate operational technologies, such as on-orbit replacement of key instrument interchangeable modules, with a focus on space robotics [1].

In 2003, the Canadian company MD Robotics proposed the “Smallsat Servicer” project for low-cost and efficient space maintenance services, in which the small satellite Smallsat is treated as a floating base and operations such as refueling, expansion of external propulsion components, and replacement of momentum wheels and controllers are carried out by multiple space robotic arms on board [49]. The project has now verified the feasibility of Smallsat docking to a spacecraft without a dedicated robotic interface using an analog simulation.

In 2005, NRL and the Naval Center for Space Technology (NCST) jointly developed the Spacecraft for the Universal Modification of Orbits (SUMO) project, shown in Figure 10. The program uses the spacecraft as a basic platform equipped with three seven-degree-of-freedom space robotic arms, and further integrates machine vision and autonomous control algorithms to lay the structural foundation for capture rendezvous and fine manipulation of the spacecraft [50].

In 2007, DARPA and Boeing successfully orbited the Orbital Express program, as shown in Figure 11. In this project, the ASTRO (Autonomous Space Transfer and Robotic Orbiter) service satellite is equipped with a space robotic arm that can independently capture and dock with the target satellite NextSat (Next generation Satellite), and disassemble, assemble, and replace the target satellite’s modules through the robotic arm system. The aim of the project is to develop a future-proof, standardized architecture for floating satellite services and to explore the validation and readiness of the technologies required for this process [51]. To date, Orbital Express has conducted autonomous component replacement, fuel replenishment, and autonomous rendezvous and docking validation, providing a solid foundation for the development of more complex on-orbit operating systems.

In 2010, GSFC launched the Notional Mission (NM) series of missions [1,17], as shown in

Table 2. Notional Mission Summary Table [17].

Case	Full Project Name	Orbit	Task/Service	Task Execution	Customer Design	Attitude
NM1	GEO Supersync	GEO	orbit modification	robots	legacy	spinning
NM2	GEO Refueling	GEO	refueling	robots	legacy	3-axis stabilized
NM3	LEO Refurbishing	LEO	upgrade	humans (COTS) + robots	designed for upgrade	3-axis stabilized
NM4	Earth-Moon L1 (EML1)	EML1	assembly	robots	designed for assembly	3-axis stabilized
	Robotic Assembly
NM5	Highly Elliptical Orbit	HEO	upgrade	humans (Orion) + robots	designed for upgrade	3-axis stabilized
	(HEO) Refurbishing
NM6	Sun-Earth L2 (SEL2)	SEL2	assembly	human (Orion + habitat)	designed for assembly	3-axis stabilized
	Human/Robotic Assembly			+ robots

. Among the NM series of missions, NM1 aims to develop a servicing spacecraft, as shown in Figure 12a, that can capture and control satellites in synchronous orbits using four 2 m-long space robotic arms and has the capability to transfer to target orbits. NM2, shown in Figure 12b, focuses on geosynchronous orbit fuel replenishment services. The aim of NM3 is to develop a dexterous service module (DSM) equipped with two robotic arms of short and long length to meet the on-orbit service needs of LEO satellites through space robotics, as shown in Figure 12c. The goal of NM4 is to develop a floating robotic telescope construction servicer (RTCS), as shown in Figure 12d, which will serve as a basic platform for the construction of a 30 m space telescope at the first Lagrangian point of the Earth-Moon by space robotics. NM5, shown in Figure 12e, is designed to provide astronaut and robotic on-orbit servicing of the Advanced Technology Large Aperture Space Telescope ATLAST-9.2m being developed by NASA. NM6 aims to develop a human–robot collaborative servicing spacecraft, as shown in Figure 12f, with the goal of constructing a manned or fully robotic on-orbit servicing platform for assembling and maintaining large telescopes at Sun-Earth L2 (SEL2).

In 2011, DARPA started the “Phoenix” program, as shown in Figure 13. The plan is to use a floating space robotic arm system to recover and reuse components or modules of abandoned satellites in geosynchronous orbit. Ultimately, this initiative will result in the assembly of entire satellites based on in-orbit recyclable materials. The specific operation is to launch a modularized “cell” at the sub-system level or component level of the satellite into the geostationary orbit, assemble it on the antenna of the abandoned satellite using the universal space robot arm of the SUMO project, disassemble it as a whole to form a new satellite, and then transfer it to the target orbit to release it. The Phoenix aims to change the traditional paradigm by developing new satellite assembly architectures, thereby reducing the cost of space-based systems [52,53].

In the same year, the Canadian company MDA, in collaboration with the International Telecommunications Satellite Organization Intelsat, developed the Space Infrastructure Servicing project, shown in Figure 14, which aims to provide on-orbit refuelling services using a space robotic arm system [54].

In 2013, GSFC proposed the MAST (Modular Assembled Space Telescope) concept for modular assembly of space telescopes, as shown in Figure 15. It realizes the assembly of the entire instrument by reassembling the 12- and 16-mirror segments assembled on the ground by astronauts and robots in low orbit, and sending them to higher orbits for deployment after completion [55,56]. As the project is still at the conceptual stage, it has not yet been decided whether it will take the form of multiple independent floating space robots or semi-fixed assembly robots.

In 2014, DARPA proposed the Robotic Servicing of Geostationary Satellites (RSGS) project, as shown in Figure 16. The RSGS project is realized in the form of a satellite-based robot that builds a dexterous multi-arm operating capability through a multi-robot system mounted on a pedestal, thus providing technical support for the restoration, maintenance and life extension of satellites in geosynchronous orbit. The implementation of this project has significantly improved the reliability of the US space infrastructure [57,58]. All component-level testing for the RSGS programme was completed in August 2022. The program has now reached several key milestones, including the completion of two dexterous robotic arms designed for inspection and maintenance and their stress testing for on-orbit environments, as well as the integration of the robotic arms with their associated electronics, tools, and ancillary hardware to form a fully integrated robotic payload.

In 2014, NASA initiated the OSAM-1 (On-orbit Servicing, Assembly, and Manufacturing 1) program, as shown in Figure 17. The objective of the project, designated “Restore-L” during its preliminary phase, is to construct robotic spacecraft that are equipped with the requisite tools, technologies, and maneuverability for on-orbit servicing. These spacecraft are designed to perform tasks such as inter-satellite rendezvous, grappling, fueling, and repositioning, which can be facilitated by a robotic arm system [59]. The OSAM-1 spacecraft is equipped with a 5 m-long SPIDER robotic arm, which is capable of performing the complex on-orbit assembly operation of assembling up to seven communication components into a communication antenna [60,61]. The project, which completed its critical technology review in February 2022, was scheduled to install three robotic arms and was expected to be launched no earlier than 2025. It was aimed to perform on-orbit servicing for the Earth Resources Satellite 7 (Landsat 7), which was launched in 1999. However, on 29 February 2024, NASA announced that it was terminating the project, in part due to ongoing technical, cost, and schedule challenges.

In 2015, NASA proposed the “Dragonfly” program, which is one of the three “Tipping Point Programs”, as shown in Figure 18. The Dragonfly is a lightweight system equipped with a dexterous 3.5 m robotic arm, which is capable of clamping and maneuvering objects from either end and can also be used for operational control [62]. The objective of the project is to study the on-orbit assembly technology with large scale advanced robotic systems. Additionally, the project is oriented towards the on-orbit precision assembly of satellite antennas and large satellites. It has now completed all ground demonstration verification experiments of the satellite’s large reflector antenna using the robotic arm system. Furthermore, the project’s robotic arm system has been renamed SPIDER (Space Infrastructure Dexterous Robot) and is employed in the OSAM-1 program [63].

In the same year, NASA initiated the SALSSA (Space Assembly of Large Structural System Architectures) program, as shown in Figure 19. The project is focused on three major categories of upgradable and reconfigurable large-scale modularized structural systems, such as large-scale space observatories, megawatt solar arrays, and the reuse of Mars mission components. The objective is to achieve automated on-orbit assembly, service assurance, refurbishment, reconfiguration, and reuse [64].

In 2016, the Commercial In-space Robotic Assembly and Services (CIRAS) program, a sub-program of NASA’s Tipping Point Program, was initiated. CIRAS is comprised of three principal components: TALISMAN (Tension Actuated in Space MANipulator), as shown in Figure 20a; SAMURAI (Strut Assembly, Manufacturing, Utility and Robotic Aid); and NINJAR (NASA Intelligent Jigging and Assembly), as depicted in Figure 20b. TALISMAN is capable of demonstrating the deployment of solar arrays, NINJAR can be used for autonomous construction of trusses in space, and SAMURAI is capable of strut assembly, manufacturing, and robotic aid [65,66]. The CIRAS project aims to develop technologies for the autonomous construction of large platforms in space [16]. Currently, the project has successfully completed ground demonstrations of the aforementioned three components, and plans to carry out on-orbit demonstrations in the second phase.

In 2016, following OSAM-1, NASA proposed the OSAM-2 program, which was also a subprogram of the Tipping Point Program. It was originally designated “Archinaut”, as shown in Figure 21. The objective of the program is to utilize robotic arms to assemble space-augmented truss beam structures into solar arrays in space [67,68]. The project has demonstrated the robotic arm’s ability to fabricate and assemble in a simulated space environment and has passed mission critical design review. Although the project concluded in 2023, it yielded invaluable insights and safeguarded crucial project data, establishing a robust foundation for the advancement and implementation of future OASR technology.

In 2019, NASA launched Astrobee, an in-cabin service robotics program with the goal of developing a high-performance robot capable of working for extended periods of time without the need for supervision or intervention by astronauts [69]. The project is designed with three cube robots, each equipped with robotic arms, named Honey, Bumble, and Queen, as shown in Figure 22. The Astrobee project’s floating robots can use robotic arms to grip cabin handrails, thereby conserving energy, or to assist astronauts in inventorying items, documenting experiments, delivering cargo, and other tasks [70]. Furthermore, the Astrobee project has devised an innovative propulsion method that employs a robotic arm for self-throwing, enabling a wider range of motion with reduced energy expenditure [71]. Currently, the Astrobee system has helped dozens of organizations with space experiments on ISS as a research platform, and it has made significant contributions to the advancement of space robotics [72].

4.1.2. Status of Floating Space Assembly in Europe

The ESA and the DLR have spearheaded the exploration and research in the field of on-orbit assembly with floating space robots. They have planned and implemented a number of diversified research projects focusing on on-orbit assembly technology and conducted a multitude of key technology simulations and experimental verifications.

Already in 1990, the ESA initiated the GSV (Geostationary Service Vehicle) program, as shown in Figure 23a. The GSV project satellites are equipped with a space robotic arm system that enables the repetitive grasping and repair of faulty satellites in geosynchronous orbit [73,74]. The main mission of GSV is to conduct contactless measurements of target satellites when approaching them to within 10 m, and to assist the recovery of malfunctioning satellites, as well as the re-orbiting of satellites. Additionally, the demonstration of the validation of rendezvous and sensing technologies is carried out through the GSV-X (GSV-eXperimental) project [75].

In 2004, DLR, EADS (European Aeronautic Defence and Space Company) and the Babakin Science and Research Space Centre jointly proposed the TECSAS (Technology Satellite for the Demonstration and Verification of Space Systems) program, as shown in Figure 23b. The project comprises two satellites. The service satellite is equipped with a seven-axis space robotic arm and gripping system [76,77], which is to validate the hardware and software requirements for key space robotic systems serving space maintenance and repair. However, the program was subsequently terminated in 2006 as DLR shifted its focus to the DEOS program.

In 2011, DLR initiated the DEOS (Deutsche Orbital Servicing) program, as shown in Figure 23c. The configuration of the DEOS project is analogous to that of TECSAS, which is also constituted of two satellites: one for the simulation of non-cooperative target satellites and the other for the performance of services.

Both satellites will demonstrate the operations required to provide future satellite services, including long-range rendezvous, non-target capture, calibration of spacecraft complexes, controlled de-orbiting of complexes, rigidly coupled satellite configurations, and so forth [78,79]. The DEOS is designed to identify and assess the procedures and technologies necessary for performing a rendezvous, capture, and subsequent disengagement from non-cooperative satellite operational orbits [80].

In 2012, DLR proposed the iBOSS (Intelligent Building Blocks for On-Orbit Satellite Servicing) project, as shown in Figure 23d. The iBOSS project is designed to investigate the potential of a reconfigurable spacecraft with standardized interfaces and a modular design, with the aim of enabling on-orbit assembly and thus reducing maintenance costs [81,82]. The objective of the second phase of the project, iBOSS-2, is to further develop the modular space system through the integration of autonomous service robots capable of performing maintenance and repairs during the operational lifetime of the system [83]. The third phase, iBOSS-3, is aimed at developing a modular and reconfigurable concept satellite that breaks down a conventional satellite bus into subsystems for integration into standardized smart modules [84]. In December 2022, the project’s iSSI (intelligent space system interface) modular coupling suite was demonstrated on the ISS orbit and tested on its external KIBO platform, among other things.

To promote the sustainable use of space, improve flexible and evolvable space system technologies, and develop on-orbit assembly and service capabilities, Thales Alenia Space, GMV Innovation Solutions, DLR, and others are launching the EROSS (European Robotic Orbital Support Services) project series. This series of projects aims to improve European autonomous and reliable on-orbit service capabilities, including satellite refueling, payload replacement, maintenance, orbit transfer, re-entry of space debris, etc. [85].

The series was implemented in multiple projects by phase, the composition and development of which is shown in Figure 24. The years 2019 to 2021 are the initial phase of the EROSS series of projects. The EROSS, which integrates and enhances the previous OG1 through OG6 programs, is part of the Strategic Research Cluster (SRC) on Space Robotics Operational Grants (OG7). These includes the ESROCOS (European Space Robot Control Operating System) project dedicated to the design of Robot Control Operating Software, the ERGO (European Robotic Goal-oriented autonomous cOntroller) project aiming at providing the most advanced but flexible space autonomous framework/system for single and/or multi-associated space robotic means/missions, the FACILITATORS (Facilities for testing orbital and surface robotics building blocks) project for high-level validation of space robots [86,87], the InFuse (Infusing Data Fusion in Space Robotics), I3DS (Integrated 3D Sensors suite), SIROM (Standard Interface for Robotic Manipulation of Payloads in Future Space Missions), and so on.

Some of these projects, such as the InFuse, SIROM, and FACILITATORS, used an on-orbit servicing simulator (OSS-SIM) developed by DLR during the experiments, as shown in Figure 25. The simulator uses two industrial robots to simulate the weightlessness of the service unit and the target satellite by grasping them, respectively, and a lightweight robotic arm is attached to the service unit to service the target satellite [89,90]. The system will continue to provide simulation experiments for various on-orbit programs, including robotic teleoperation, capture of non-cooperative tumbling target satellites, and space debris removal.

The EROSS+ project from 2021 to 2023 is Phases A and B1, with the goal of designing a demonstration setup while advancing the maturity of key building blocks [91]. The project demonstrated the autonomous grasping and manipulation required for the mission on the ground using a CAESAR robot, thereby validating the basic technologies, including sensors, actuators, algorithms, and software frameworks, as shown in Figure 26a. The EROSS SC project from 2023 to 2025 is phases B2 and C, with the objective of continuing to enhance the maturity of the technology in order to achieve all the functionalities before the on-orbit demonstration in 2026, as illustrated in Figure 26b. Subsequently, the series of projects will continue, with demonstration operations such as docking, refueling, and ORU replacement planned for low Earth orbit in 2026, on-orbit service operations in geosynchronous Earth orbit in 2027–2028, and full realization of autonomous assembly missions in orbit after 2035.

4.1.3. Status of Floating Space Assembly in Japan

Japan’s research in the field of on-orbit assembly with floating space robots is best known for the ETS series projects that have been carried out successively since 1975. Among these, the ETS-VII satellite, launched in 1997, was the first known satellite to be equipped with a space robotic arm system, as shown in Figure 27. The space robotic arm carried on ETS-VII not only demonstrated the satellite capture, rendezvous, and docking functions, but also verified the key on-orbit assembly technologies, such as truss structure and test antenna assembly [93].

4.2. Status of Research on Semi-Fixed Space Assembly

One of the primary forms of OASR is the semi-fixed space assembly with robots. This configuration effectively combines the high stability, flexibility, extensive range of motion, and robust task adaptability of space robotics.

4.2.1. Status of Semi-Fixed Assembly in North America

The research and exploration of technology for semi-fixed space assembly with robots in North America is primarily conducted by the United States. Key institutions engaged in this research include NASA, JPL, and the Goddard Space Center, as well as universities and military industrial enterprises that collaborate with these institutions. These entities have been engaged in research and exploration of related technology for a longer period and have amassed a comprehensive and extensive repository of experience.

In 1999, NASA and Carnegie Mellon University collaborated on the development of the Skyworker space robotics system, as shown in Figure 28. The Skyworker system was designed to validate and demonstrate key technologies for space assembly, inspection, and maintenance. The robotic arm system is capable of autonomous movement on a truss structure, with a range of up to several kilometers, while the load-moving capacity is up to several tons. This enables the system to complete on-orbit assembly, inspection, and maintenance tasks for large space structures, such as space solar power stations [94,95].

In 2006, the Goddard Space Center in the United States proposed a conceptual project for the autonomous assembly of a 30 m-class space telescope using a space robotics system, as shown in Figure 29. The conceptual project proposes an entirely robotic telescope assembly process, with astronauts acting as supervisors only when necessary [7]. The assembly robot is an 18 m multi-arm system consisting of seven-degree-of-freedom robotic arms, which can alternately travel on the telescope structure and assemble the optics with the necessary precision.

In 2012, NASA proposed the famous SpiderFab space manufacturing system concept, as shown in Figure 30. One of the key characteristics of the SpiderFab system is its semi-fixed space assembly configuration, which employs a spider-like multi-arm robotic system capable of self-fixation on the assembled structure. This configuration is designed to address the unique challenges of on-orbit additive manufacturing and robotic cooperative assembly. It enables the on-orbit manufacturing of multi-functional structural components and the on-orbit assembly of such large space structures, such as antennas, solar panels, and trusses, among other tasks [96]. The project has thus far yielded the development of robot tools and control software, as well as the completion of experimental verification of ground assembly and connection of trusses [97].

In 2016, the JPL and the California Institute of Technology jointly proposed the RAMST (Robotically Assembled, Modular Space Telescope) project, as shown in Figure 31. The project involves the modular assembly of a 100 m-class aperture very large space telescope by means of a six-legged robot, designated “Hexbot”, which uses a robotic system to assemble and connect all trusses and then installs mirror modules by crawling and moving on the trusses [98,99].

In 2019, the NASA Langley Research Center initiated the Assemblers program. It is based on the NINJAR system in the CIRAS project and fully integrates the concept of modular reconfigurable robot design. This allows for the development of a robot system that can be reconfigured into an optimally matched structure according to the specific assembly task objectives, as shown in Figure 32. The robotic assembly system consists of a stack of modular platforms, each with a varying number of actuators and six degrees of freedom of motion between the platform bases. The platforms are equipped with component position feedback sensors. The program has completed the initial construction of a prototype and performed ground testing of a multi-layer stack of components. Plans for future use of the system include space targets or facility construction missions on the surface of an exoplanet [100].

In 2017, NASA’s Ames Research Center introduced the Automated Reconfigurable Mission Adaptive Digital Assembly Systems (ARMADAS) program, as shown in Figure 33a. The goal is to investigate the next generation of infrastructure, instrumentation, and autonomous assembly and manufacturing capabilities for spacecraft in response to the needs of exploratory science missions [101,102]. The project envisions the autonomous assembly and construction of a variety of large-scale structures, including space telescopes, solar arrays, towers, and other infrastructure, using ultra-lightweight, high-performance mechanical metamaterial modular units via a robotic arm system called SOLLE (Scaling Omnidirectional Lattice Locomoting Explorer), which is divided into external transport robot and the internal fastening robot [103]. The project has now completed the verification of the autonomous assembly of a large number of structural modules in a terrestrial gravity environment. The results demonstrate that the mechanical properties of the structures, as well as the robotic assembly performance, exhibit excellent accuracy and cost-effectiveness. Additionally, the project has conceptualized a vision for the construction of a lunar surface infrastructure, as shown in Figure 33b.

4.2.2. Status of Semi-Fixed Assembly in Europe

There is a paucity of systematic research projects in Europe pertaining to the domain of semi-fixed space assembly with robots. The British company Frazer-Nash had proposed a project for the construction of a kilometer-scale space-based solar power plant in 2020 [105]. Its satellite module assembly is planned to be carried out by autonomous robots in medium Earth orbit (MEO) to avoid the effects of space debris in low Earth orbit (LEO) and the harsh radiation environment of the inner Van Allen belts. The project has now been demonstrated to be feasible and the first phase of its target implementation has commenced in 2022. The second phase, scheduled to take place between 2027 and 2031, will involve the design of the assembly robotic system.

4.2.3. Status of Semi-Fixed Assembly in China

China has also proposed its own project in the field of semi-fixed space assembly with robots. In 2014, the Changchun Institute of Optics, Fine Mechanics and Physics (CIOMP) of the Chinese Academy of Sciences proposed a plan for the construction of an ultra-large aperture space telescope, which may employ a space robotics system to assist in the on-orbit assembly of the telescope [106], as illustrated in Figure 34. The project has now developed three generations of super-redundant robotic arms, all of which are spliced together by nine modular joints. Among these, ground-based splicing experiments for large telescopes have been completed using the third generation of super-redundant robotic arms.

In 2022, the HIT proposed a scalable self-attaching/-assembling robotic cluster system (S2A2RC), as illustrated in Figure 35. The system is capable of accomplishing the on-orbit assembly of ultra-large-scale space truss structures through the use of multiple cluster robots. Each robot crawls on the truss through a hexapod mobile mechanism, and the operating system consists of a five-degree-of-freedom robotic arm and a gripper jaw. An electrostatic sensor is integrated on this robot to monitor the crawling, gripping, and assembly status of the robot in real time during the assembly process, thereby enhancing the reliability of the assembly [107,108].

4.3. Status of Research on Fixed Space Assembly

The development and research of fixed assembly with space robots started earlier, and most of the technology verification in the early stage was done with the help of the robotic arm system on the ISS, such as SSRMS, SPDM, JEMRMS, CSSRMS, and so on. Although there are relatively few systematic research projects, it is one of the important components of the OASR technologies. Typical stationary on-orbit assembly programs include NASA’s Robotic Refueling Mission (RRM), DLR’s Space Factory 4.0, and AIRBUS’ PERASPERA In-Orbit Demonstration (PERIOD).

The RRM is a robotic space fueling program initiated by NASA in 2011, as shown in Figure 36. The aim of the project is to evaluate the efficacy of tools and associated key technologies for a robotic arm to perform fuel refuelling services in orbit, particularly for spacecraft lacking a dedicated fuel refuelling design. The RRM1 and RRM2 phases of the project have been trialled on the ISS, with a series of demonstrations conducted with the assistance of SPDM [109]. In the RRM3 phase, NASA made a significant advancement in the technology for storing and transferring cryogenic fuels in space [110].

Space Factory 4.0 is a project undertaken by DLR and the German Federal Ministry of Economics and Energy (BMWi) in 2017, as shown in Figure 37. The project aims to investigate the new processes and technologies of an on-orbit platform for the rapid assembly of space robotic satellites. This objective can be achieved through the use of digital twin technology and other support systems for the on-orbit assembly and manufacturing of highly modular satellites [111]. The project was designed and developed with bilateral controllers that allow a human operator to remotely operate the assembly robot using a human–machine interface, while providing highly realistic stress feedback to the operator with the support of a virtual fixture. The addition of the teleoperation feature also provides the space robotics system with more flexibility and task adaptability.

The PERIOD is a research program initiated by Airbus in 2020, as shown in Figure 38. Its objective is to develop the means for on-orbit construction of functional satellites, with the assistance of a space robotic orbital factory. In essence, the PERIOD represents a space processing and manufacturing platform working in orbit around the Earth. This platform is equipped with a system of space robotic arms, which are deployed to carry out the on-orbit production of parts and modules for spacecraft, including satellites. Additionally, the system is capable of constructing antenna reflectors, assembling spacecraft components, and replacing satellite payloads. This paves the way for the future on-orbit manufacture of large-scale structures [112,113].

4.4. Summary of the Space Robot On-Orbit Assembly Project

This section summarizes the on-orbit assembly projects carried out by the world’s major space agencies that were investigated. The final results are presented in

Table 3. Summary of projects for on-orbit assembly with space robots.

No.	Forms	Region	Year	Agency	Project Name	On-Orbit Goals	Number of Robotic Arms
1	Floating	North America	1999	Boeing	NNGST	Space telescope	Concept
2			2003	Boeing	AAST	Space telescope	Concept
3			2003	AFRL, SMC, NRL	XSS	Inspection, docking, servicing	XSS-12: Concept
4			2003	Canada MD Robotics	Smallsat servicer	Maintenance, refueling	1
5			2005	NCST	SUMO	Satellites, debris	3
6			2006	GSFC	TMST	Space telescope	4
7			2007	DARPA, Boeing	Orbital Express	Satellite capture, parts replacement	1
8			2010	GSFC	NM	NM1: Non-cooperative satellite capture, re-orbiting;	NM1: 4;
						NM2: Fuel filling;	NM2: 2;
						NM3: On-orbit services;	NM3: 2;
						NM4: Space telescope construction;	NM4: 2;
						NM5: 2 Arms Capture and Berthing + 2 Arms Operation;	NM5: 4;
						NM6: Space telescope construction	NM6: 4;
9			2011	MDA	SIS	Refueling	1
10			2011	DARPA	Phoenix	Dis/Assembly, maintenance, Fuel filling	4
11			2013	GSFC	MAST	Space telescope	Concept
12			2014	DARPA	RSGS	Satellite maintenance and repairs	2
13			2014	NASA	OSAM-1	Assembly, service	3
14			2015	NASA	Dragonfly	Assembly	1
15			2015	NASA	SALSSA	Assembly, repair, reconfiguration	Concept
16			2016	JPL	RAMST	Space telescope	6
17			2016	NASA	CIRAS	Trusses, reflectors, satellites	2
18			2016	NASA	OSAM-2	Solar array assembly	1
19			2019	NASA	Astrobee	Services within ISS	1
20			2019	TAS, GMV, DLR	EROSS series	Orbital support services	1
21		Europe	1996	ESA	GSV	Satellite capture, re-orbiting	1
22			2004	DLR, EADS, Babakin	TECSAS	Satellite maintenance and repairs	1
23			2011	DLR	DEOS	On-orbit services	1
24			2010	DLR	iBOSS	Modular spacecraft	2
25		Japan	1997	JAXA	ETS-VII	On-orbit services	1
26	Semi-fixed	North America	1999	NASA, CMU	Skyworker	Assembly and maintenance of large space structures	Concept
27			2012	NASA	SpiderFab	Additive manufacturing and assembly of large space structures	4
28			2019	NASA, LaRC	Assemblers	Planetary surface infrastructure construction	Reconfigurable
29			2017	NASA	ARMADAS	Construction of large space structures	1 (Multi-robots possible)
30		Europe	2020	Frazer-Nash	Space-Based Solar Power	Space solar power station	Concept
31		China	2014	CIOMP	Ex-Large Aperture Space Telescope	Space telescope	4
32		China	2022	HIT	S2A2RC	ultra-large-scale space truss	7 (Multi-robots possible)
33	Fixed	North America	2011	NASA	RRM	Fuel filling	2
34		Europe	2017	DLR, BMWi	Space Factory 4.0	On-orbit manufacturing	Concept
35			2020	AIRBUS	PERIOD	Space in-cabin assembly	1 (Multi-robots possible)

. In particular, a statistical graph of the number of robotic arms utilized in space assembly projects for non-large size space equipment or facilities (i.e., regular size spacecraft) is shown in Figure 39.

5. Literature Analysis of On-Orbit Assembly with Space Robots

In this section, the research hotspots and development trends in this field are sorted out based on the results of the statistical analysis of the related papers about the OASR. As the basic method of research analysis, bibliometric analysis aims to accurately mine the hotspots in the field from the academic research dimension as well as sort out the distribution of research power and the trajectory of technological evolution. Relying on the comprehensiveness and objectivity of literature data, this method provides quantitative support for identifying key technology clusters, revealing international research networks and development trends, and is an important link between preliminary classification research and subsequent technical analysis. First, the relevant literature was searched using Web of Science, and the search expression used was “(assembly OR manufactur *) AND ((in space) OR (on orbit)) AND (robot)” and the search time range was 2014∼2024, with 1322 relevant papers. Then, the CiteSpace is used to visualize and analyze the retrieved literature and to sort the analysis results. The literature analysis related to OASR focuses on both the comparison of the number of published papers by country or institution and the trend of research hotspots over time, and the results of the analysis are shown below.

First, the retrieved papers were clustered and analyzed by country using CiteSpace to roughly compare the research contributions of each country to the field during 2014∼2024, and the results are shown in Figure 40. In Figure 40, the size of the circle indicates the total number of papers published in the field by each country, the inner ring indicates the number of papers per year, the thickness of the ring indicates the number of articles published in the corresponding year, and the outermost red ring represents the latest year. It can be seen that the countries with more active research in this field are China, the United States, Germany, England, Australia, and so on. Although China started late in this field, it has become more active in recent years.

The most active institutions in the world in the research of OASR, as shown in Figure 41, are: NASA, University of Western Australia, HIT, Chinese Academy of Sciences (CAS), and so on. Among them, NASA, as the central coordinating body for relevant research institutions, initiated research into the OASR at an earlier stage. This has resulted in a significant accumulation of technical expertise and technological resources. In China, institutions such as the CAS and the HIT have played a significant role in the country’s accelerated advancement in this field.

The trend of research hotspots towards OASR over time, derived from the keyword clustering analysis of the literature, is shown in Figure 42. The time zone in which each keyword node in the figure is located indicates the year in which the keyword first appeared, and the more frequently the keyword appeared later, the larger the node is. The rows represent distinct keyword clusters, or research directions, which are labeled on the right side of each row. It is a refinement and generalization of all the keywords in that row, making it convenient to understand the different research directions in this field. The connecting lines between nodes indicate the relationship between keywords. The length of the connecting lines reflects the temporal span of research activities within a given direction. As illustrated in the figure, the research hotspots in this area mainly include robotic assembly and related methods, space robot, digital twin technology, etc. Of these, motion control of space robots and task planning have emerged as the most prominent and enduring research topics in recent years.

Preliminary conclusions drawn from the aforementioned analysis include the following: (1) With respect to the exploration of space technology, research on OASR in various countries has been gradually active. (2) The analysis reveals relevant research is usually dominated by major institutions, accompanied by an observable trend towards networking and decentralization, with these dominant institutions maintaining a central role. (3) The design of space robots is relatively mature by default, and the progress of the structural design of space robots with new structures, forms, and driving principles in orbit has slowed down. (4) The initial phase is characterized by conceptualization and design, with subsequent advancements in the field of artificial intelligence gradually becoming a significant driving force for the development of space assembly technology on orbit. Concurrently, there is an increasing inclination towards the verification of specific functions, indicating the OASR-related technologies are gradually accumulating maturity. (5) The future may be marked by the integration of space-applicable artificial intelligence, innovative structural designs, and systematic platform design technologies.

6. Key Technologies for On-Orbit Assembly with Space Robots

While space robotics on-orbit assembly technology offers considerable advantages and represents a crucial future development trend, it also presents a number of challenges and is still a long way from large-scale practical deployment and application. Firstly, the influence of the space environment, comprising microgravity, a wide temperature range, strong radiation, and a high vacuum on the process of multi-robot on-orbit assembly cannot be ignored. The harsh environmental conditions of space operations put forward higher requirements on the reliability, stability, and adaptability of space-based platforms, space robotic systems, and operational targets. Secondly, the remote or ultra-remote deployment of space robotic on-orbit assembly systems also presents communication challenges, including high latency, strong interference, and low bandwidth for ground interactions. Finally, on-orbit assembly of space robots represents a multidisciplinary integration of systems engineering. The development of effective strategies for ensuring stability, efficiency, intelligence, and autonomy in these operations remains a significant research challenge. For these reasons, this section presents an overview of some principal technologies involved, as illustrated in Figure 43.

6.1. Ground-Based Simulation Technology

Ground simulation technology represents a significant avenue for the validation of the theoretical and methodological frameworks underpinning the robot on-orbit assembly. Firstly, the microgravity environment in space allows both the robot and the object to be assembled to float, which results in the control of the robot’s joint torque differing from that on the ground. Furthermore, the process of assembly may present difficulties in locating and controlling the attitude of the target object, and non-target objects in a floating state within the assembly environment may pose a significant risk to the safety of the robot’s motion. Moreover, the robust radiation environment and notable light temperature disparity in space will exert a more pronounced influence on the robot’s observation system, remote control system, and overall operational lifespan. In comparison to the terrestrial environment, the light intensity in space is approximately six times that of the ground, with a temperature differential between light and shadow surfaces reaching upwards of 300 °C. The oscillating light and darkness resulting from the operational orbit necessitates a heightened degree of precision and dependability from the observation system. In addition, the high vacuum space environment necessitates adaptations to the structure of the robotic system. Thus, the simulation of the complex space environment on the ground is a crucial method for enhancing the efficiency of spacecraft or space robot development, particularly in the context of on-orbit assembly process verification.

In order to emulate the conditions of a space mission, various countries have created dedicated simulation devices with specific objectives and mission-oriented applications. NASA’s Neil Armstrong Test Facility (ATF), which encompasses the Space Environments Complex (SEC), the Combined Effects Chamber (CEC), and other space environment simulation apparatus, is illustrated in Figure 44. The system is capable of simulating a variety of special environments, including vacuum, low temperature, acoustic interference, complex vibration, electromagnetic interference, and others. This enables the conduct of ground tests for corresponding U.S. space missions [114,115]. Furthermore, NASA’s Zero Gravity Research Facility (ZGF), the C-9 low-gravity flight research aircraft (also known as the “Vomit Comet”), and other analogous devices have also contributed to the advancement of microgravity simulation technology in space [116,117].

In light of the accelerated pace of space exploration in China, the demand for ground simulation technology for space environments is equally urgent. The Harbin Institute of Technology (HIT) spearheaded the development of the “space environment ground simulation device”, as shown in Figure 45. This device is a comprehensive space environment simulation laboratory that can effectively simulate a multitude of space environmental factors, including microgravity, vacuum, high and low temperature, charged particles, electromagnetic radiation, space dust, plasma, weak magnetic field, neutral gas, and other nine categories. It is the world’s largest, most abundant, and most comprehensive space environment simulation facility, which has significantly enhanced humans’ capabilities in on-orbit assembly, innovation, and development in the fields of deep space exploration and even extraterrestrial exploration and development [118]. The utilization of a large-scale ground air suspension apparatus and a six-degree-of-freedom microgravity simulation device also serves as an efficacious supplement for the ground microgravity simulation of space robots [119]. Furthermore, the 207 Institute of the Second Academy of Aerospace Science and Industry of China has developed the largest vacuum solar simulator in China, which is capable of effectively simulating sunlight in space. This has the potential to facilitate the advancement of the robot’s capacity to measure and sense the environment of on-orbit assembly [120].

Ground simulation has different focuses for different forms of assembly tasks. In the context of the fixed space assembly, the primary focus of ground simulation is on verifying the adaptability of space robots in the space environment, their dexterous operation and collaboration ability, and high-precision control. For the semi-fixed space assembly, the ground simulation must prioritize stability and positioning accuracy during finite-degree-of-freedom movements. Additionally, it is essential to assess the adaptability of the robot when operating in varied attitudes, such as during mobility assembly tasks within the confines of a space truss movement. The ground simulation of floating space assembly is an even more arduous task, necessitating an effective simulation of the motion control of the robot in the free-floating state, as well as the docking control with the target object, to cope with the complex assembly conditions in the actual space.

In view of the sustained high cost of space launches, the advancement of space environment ground simulation technology and the construction of associated devices provides an important boost for addressing the challenge of accurately simulating the environment of robots on-orbit assembly. Nevertheless, a significant discrepancy persists between ground simulation and the actual space environment. The development of a comprehensive simulation technology encompassing multiple environmental factors, such as microgravity, vacuum, temperature extremes, illumination, electromagnetic interference, and others, remains a formidable challenge.

6.2. Space Robot Morphology Theory and Technology

In the context of the increasing complexity of spacecraft structures and functions, coupled with the diversification of space missions and objectives, there is a compelling necessity for the advancement of more sophisticated space robotic systems to support these developments. The objective of space robotics morphology, which is concerned with the study of the organization, structure, design, and functionality of robotic systems, is to facilitate more efficient performance of specific tasks or functions by designing appropriate organizational and structural forms for robotic systems.

In order to achieve optimal assembly outcomes in diverse operational environments, it is essential to analyze the suitability of various organizational structures [121], including cluster robotic systems and heterogeneous single-unit robotic systems, for both large-scale structures, such as space telescopes and space solar power stations, and small-scale components, such as antennas and solar panels. Furthermore, it is imperative to investigate the design of robot configurations oriented to the needs of the assembly tasks, such as modular robots, multi-arm robots, soft robots, and end-effectors, with the objective of optimizing and facilitating their operability, accuracy, flexibility, and coordination. The aforementioned is a theoretical and design basis of the realization of the OASR technology.

As a modular robot, it is capable of adapting its configuration through the reconstruction of the connection relationship between modules, thereby ensuring its suitability for space operations within a dynamic mission environment [122]. Illustrative examples of this adaptability include the SuperBot robot, a collaborative effort between the University of Southern California and NASA [123], and a reconfigurable modular space robot developed by Northwestern Polytechnical University (NPU) [124], as depicted in Figure 46a and Figure 46b, respectively.

The multi-arm space robot is endowed with the capacity to crawl and move. During the assembly process, some of the robotic arms can be employed as fixed feet, while the remaining robotic arms are utilized to undertake operations such as fine assembly. This approach allows for the exploitation of the advantages inherent to a wide range of motion, a stable assembly state, and flexible operation. The semi-fixed assembly forms exhibit enhanced task adaptability, particularly in the context of large-scale space structure assembly and construction.

As for robots that possess infinite degrees of freedom, remarkable shape adaptability, and flexibility, there are irreplaceable advantages for fine manipulation in small spaces. As shown in Figure 47, the three-cable driven space grasping robot developed by HIT is able to perform three-arm or single-arm space flexible grasping tasks [125]. As these capabilities continue to evolve, it is anticipated that it will become a more effective component of the on-orbit assembly system, particularly in terms of strength, lightweight design, impact resistance, operational stability, perception, autonomous control, and other properties [126]. Moreover, its application in semi-fixed or floating assembly tasks can yield optimal performance.

The design and manufacture of end-effectors is the fundamental aspect that determines the adaptability of space robots [40]. The use of specialized actuators can significantly streamline otherwise intricate operations, as evidenced by examples such as the Grapple fixtures for walking on the space station’s surface developed by ERA [27] and the RRM project’s four replaceable end-effectors for refueling (Tertiary cap adapter, T-valve adapter, Ambient cap adapter, and Plug manipulator adapter) [127], illustrated in Figure 48. The utilization of specialized actuators has the effect of simplifying operational processes while simultaneously reducing the adaptability of the tasks in question. In contrast, a “universal actuator” that can operate other tools must consider the assembly of diverse objects and non-cooperative targets, typically exhibiting a high degree of freedom, extensive application range, and high flexibility. Examples include the double gripper of the GITAI S1, the human-like flexible hand of the Tiangong-2 robotic arm, and the multi-segment soft end designed by Harvard University [128]. However, these actuators also have inherent limitations, necessitating more complex operation logic or steps to complete a task.

6.3. Assembly Connection Structure Design Technology

In the design and implementation of future space assembly target structures, a significant trend will be the simple and unified assembly and connection of space equipment and facilities. As a pivotal strategy for structural design and component assembly, the modular concept can streamline the assembly process, enhance efficiency, and significantly reduce the burden of subsequent testing and debugging.

In the structural design of space assemblies, the replaceability, maintainability, and ease of assembly of the components are of paramount importance. In addition to the assembly and connection of the structural form, the energy supply, information transmission, transmission control, and other issues between the functional modules of each equipment are also fundamental elements of the composition of equipment and facilities. Therefore, it is particularly important to construct a uniform interface form that meets the requirements of both electrical and mechanical connection characteristics. In regard to the electrical interface design of space equipment and facilities, the current typical examples include SIROM-A, SIROM-B, electromagnetic docking interface (EDI), universal docking port, ASSIST interface, and so forth [129]. These are illustrated in Figure 49. The modular mechanical docking interface, which is based on the same composition structure and employs the same assembly locking mechanism, represents a significant means of achieving space assembly and connection.

As illustrated in Figure 50a, the mosaic matching lock button connection structure, developed by ESA in 2016, enables the robot to assemble a large space truss, which is then docked through the truss node joint and the interface at both ends of the truss rod. Similarly, a truss member and ball joint fixed structure has been developed by Huazhong University of Science and Technology [130,131], as illustrated in Figure 50b.

These modular assembly and connection components are compact, straightforward to install and commission, and are optimal for the assembly with space robots. In recent years, there has been a notable shift towards standardization and unification of module assemblies and interfaces, as well as an increasing focus on the intellectualization of modular connection plug-and-play functions. This has emerged as a pivotal technical development trend in the design and connection of space robot assembly structures.

6.4. On-Orbit Assembly Interactive Management Technology

The objective of robot on-orbit interaction management technology is to coordinate, control, and manage the constituent subsystems on the ground and in space. This is achieved by utilizing the space assembly mission environment as the initial condition, which encompasses digital twin technology, ground teleoperation technology, assembly planning and management technology, autonomous decision-making technology, and other related supporting technologies.

Digital twin technology can provide multi-dimensional and full-cycle engineering capability for on-orbit assembly. Through high-fidelity modeling, the physical coupling analysis model of the space environment can be constructed to study its impact on the performance of space robots, and at the same time, the space assembly process can be simulated to avoid safety risks in advance [132]. By two-way data synchronization of the digital twin, the assembly process can be dynamically adjusted with the help of real-time monitoring data, which can be used not only for anomaly detection but also for failure analysis and prediction in the space environment and life assessment of space robots, laying the data foundation. In addition, the application of digital twin technology can also support the application and development of other interaction management technologies. Furthermore, factors such as the microgravity of space, drastically changing light and temperature conditions, and intense space radiation have introduced significant obstacles into the field of high-precision measurement and environmental perception. The primary difficulties pertain to the mitigation of the impact of complex illumination and the differentiation and extraction of targets from the background. In particular, accurately identifying and localizing non-cooperative targets, as well as measuring their attitudes, remains a significant challenge.

Teleoperation technology represents a crucial supplementary means of realizing the interaction and control of space robots in the context of on-orbit assembly. It enables operators to intervene and assume control of the space robot system, thereby integrating human knowledge, experience, and decision-making abilities into the control loop. This, in turn, enhances the fault tolerance and adaptability of the space robot assembly system. As space exploration progresses towards deeper space, the necessity for autonomous systems is becoming increasingly apparent. Consequently, there is a pressing need to integrate a multitude of sensory data, including visual, tactile, auditory, force, temperature, and other information, in order to develop more intelligent applications [133,134]. Additionally, the limitations posed by distance and other factors necessitate the development of supporting technologies, such as delay compensation for long-distance information interaction.

Efficient process planning and operation management are fundamental to guaranteeing the orderly, coordinated, and stable on-orbit assembly system utilising space robots [135]. The space assembly process encompasses the utilization of materials, actuators, and the dynamic interaction between the two, encompassing both abstract tasks and concrete control of the robotic system. Accordingly, the process planning encompasses a series of upper-level problems [136], including task allocation, task decomposition, sequence planning, and resource scheduling, among others. Furthermore, additional considerations must be made regarding quality inspection, efficiency evaluation, self-inspection optimization, and fault handling. These factors are essential for ensuring the quality of the assembly target and facilitating the repair and maintenance of the robot itself.

6.5. Multi-Robot On-Orbit Assembly Platform Technology

Multi-robot on-orbit assembly platforms can be considered an important means of realizing on-orbit assembly, manufacturing, repair, and maintenance of space facilities and equipment. An appropriate platform design can diminish the financial burden of overseeing and operating an on-orbit production facility throughout its lifespan, considerably augment its capacity for sustainable operations, and more effectively advance the objective of human space exploration. Nevertheless, the construction of a multi-robot in-orbit assembly platform presents a multitude of considerations.

A reasonable multi-robot space-based assembly and manufacturing platform must possess the capacity to store and transfer materials in an intelligent and orderly fashion. The production and manufacturing of products entails the use of numerous materials and supporting components, including energy, tools, materials, and various parts, which are utilized in the production and manufacturing of products with diverse functionalities. Consequently, the intelligent and flexible orderly allocation of production materials is of paramount importance, as is the assurance of safety and reliability throughout the storage and transfer process.

The assembly and manufacturing process on the platform requires the implementation of a robot work system that is more flexible. The operational mode of the space-based platform differs from that of ground-based manufacturing. Such a system is distinguished by a greater proclivity towards versatility, scalability, and diversity of functions. Accordingly, the spatial configuration and operational capacity of the robot production unit are contingent upon the input task form and the anticipated outcome. For example, the assembly of spacecraft of varying dimensions requires the ability to adapt to accommodate disparate operational spaces. This necessitates the capacity to adapt the spatial configuration between units and the corresponding functional end effectors, as well as the coordination of different forms of space assembly, such as fixed, semi-fixed, and floating, in accordance with mission-specific requirements [137].

Additionally, it is also imperative to consider the external interaction capability of the space-based platform, given its role as the infrastructure for on-orbit assembly. In order to ensure the supply of space-based platform materials and provide on-orbit services to the outside world, space-based platforms must be capable of handling a significant number of spacecraft. It is therefore evident that spacecraft automatic docking and parking technology, target acquisition and attitude control technology, high-reliability locking technology, gas/liquid/circuit rapid connection or disconnection technology, and other external interaction technologies are indispensable [138]. The aforementioned technologies, such as the SUMO and orbit express projects, serve to verify relevant technologies and establish a foundation for the realization of spacecraft unmanned autonomous docking and capture operations. Nevertheless, the issue of how to manage the simultaneous berthing of a large number of spacecraft in the future remains unresolved.

7. Development Trend of On-Orbit Assembly with Robots

In consideration of the aforementioned research status and key technology analysis of robot on-orbit assembly, this section presents a summary of potential future development trend and typical applications within this field. These are primarily reflected in the following areas:

7.1. Space In-Situ Assembly

The centralized distribution of resources in space, relative to its vast range, results in significantly reduced efficiency in the acquisition and transportation of resources. Consequently, the development of technologies that facilitate the utilization of resources in situ represents a crucial area of focus for future advancements in space technology. In consideration of the aforementioned characteristics, the utilization of space robotics for the exploration, mining, extraction, and processing of extraterrestrial resources, followed by the concentration of component parts into a designated space region and the subsequent assembly of these components in situ, will constitute a primary method for the fabrication and assembly of equipment for future space facilities. For example, the “Phoenix” program of the United States represents an effort to repurpose functional components of discarded spacecraft in space and investigate the potential of utilizing these components to construct and produce new spacecraft. It is a prospective exploration of future on-orbit assembly development. Nevertheless, the majority of resources in space are still concentrated in planets or asteroid belts. The development of an in situ assembly and manufacturing system represents a core technical challenge that must be addressed in the future.

7.2. Collaborative Multi-Robot Clusters

The construction of large-scale space facilities and equipment may be best accomplished through the use of fixed, semi-fixed, and floating space robot forms or their groups, collectively known as the cluster space assembly form. The cluster operation mode allows for the collaboration of numerous operational units [139], thereby facilitating the construction of a spatial multi-robot system with diverse capabilities, high efficiency, and superior quality [140]. Furthermore, the cluster robot system exhibits enhanced robustness, fault tolerance, and flexibility, enabling adaptation to the evolving dynamics of the space assembly task environment through systematic cluster adjustment. This offers significant potential for future applications. For instance, a large number of small, independently operable floating swarm robots could be employed as a temporary positioning and auxiliary propulsion system for space-based platforms or large space structures, facilitating the delivery of targets to designated orbits or designated assembly postures, which usually involving technologies such as modular autonomous assembly, dynamic target capture [141], and distributed task assignment.

7.3. Evolution of Individual Intelligence

The evolution of the intelligence of space robots and their counterparts plays an instrumental role in determining the extent of their deployment. To illustrate, in the case of the numerous robotic execution units, in addition to global control through the central brain, it is also necessary for them to possess the capacity to interact with the surrounding environment. This enables them to achieve autonomous learning and generate intelligent images from the initial viewpoint. The operation, management, and maintenance of various equipment and facilities in the future space environment will be inextricably linked to the intelligent core that specializes in the current equipment and facilities. Accordingly, the advancement of intelligent technology to enhance the capabilities of space equipment and facilities represents a significant trend in the realm of future on-orbit assembly. Its deployment and application can establish the theoretical and technological basis for achieving fully autonomous decision-making and operation of future space robots, thereby reducing the degree of dependence on human labor. For example, Timmermann et al. [142] proposed an AI framework based on Seq2Seq encoder-decoder to achieve safe and feasible satellite module assembly sequence generation, providing a new paradigm for autonomous planning methods of complex spatial structures. This represents not only an important aspect of the advancement of artificial intelligence but also a crucial pathway for the evolution of space robotics into a more intelligent domain.

7.4. Smart Materials and Self-Assembly

The future self-assembly of space facility equipment using smart materials represents a technology with significant potential for revolutionary change in the fields of space construction and exploration. The intrinsic adaptability of smart materials enables them to modify their structure and performance in response to environmental changes and mission requirements, thereby enhancing the adaptability and survivability of space facilities. Concurrently, self-assembly technology can facilitate the automatic assembly and connection of equipment components, thereby reducing the necessity for manual intervention and consequently lowering construction costs and risks. The utilization of these materials permits the autonomous assembly and realignment of space facility equipment, thereby enhancing the efficiency and flexibility of the construction process. This technology can be applied to the construction and maintenance of space stations, satellites, probes, and other facilities, thereby providing more efficient, flexible, and sustainable solutions for space missions. This will facilitate the accelerated development of space construction, advance human exploration of distant space, and offer novel possibilities and avenues for advancement within the space field.

7.5. On-Orbit Assembly of Micro- and Macro-Scales

The field of space assembly is poised to witness a shift towards a synthesis of micro- and macro-scale developments. In the domain of micro- and nanoscale space assembly, the advent of sophisticated technologies such as nanotechnology and micro-robotics will facilitate the meticulous integration and operation of minuscule devices and structures. The utilization of micro- and nanoscale assembly technologies has the potential to facilitate the production of microsatellites, nanosatellites, and other micro-space devices, thereby enhancing their performance and functionality. Furthermore, the implementation of nanorobotics technology will facilitate the growth-oriented construction of space structures from a technical standpoint. Moreover, this technology can be utilized for the repair and maintenance of space-based infrastructure, facilitating minimally invasive repair and renewal of space facilities. On the macro scale, future space assembly technologies will be required to address the challenge of larger scale and more complex structures. This encompasses the construction and assembly of expansive space facilities, including space stations, moon bases, and Mars colonies. The use of large robots, self-assembling structures, and advanced materials will facilitate the efficient construction and operation of macro-giant-scale space facilities, which will provide the necessary support for long-term space habitation and deep space exploration. In summary, the advancement of micro- and macro-scale space assembly technologies is mutually reinforcing, collectively propelling progress in the domain of space exploration. The advancement of micro- and nanoscale technologies has facilitated the development of more flexible, sophisticated, and efficient solutions for space missions. Conversely, the evolution of macro- and megascale technologies has established a robust foundation for space exploration. These technological developments are poised to expand the frontiers of human development in space and unlock new avenues for space exploration and utilization in the future.

7.6. Space-Based Supporting Platforms

The gradual development of on-orbit assembly technology for space robots and the acceleration of human space exploration in the future will require the construction of large-scale space-based platforms, which will serve as vital nodes and springboards for advancing from Earth to deep space. As a significant component of the future program, it must encompass the assembly and manufacturing of space equipment, the repair and maintenance of spacecraft, the replenishment of space fuel, the storage of space energy, the exploration of astronomical telescopes, and other production and service functions. Additionally, as an on-orbit service platform, it must possess the capacity to undertake production and service tasks within the designated orbit or at any predetermined location in space. As a pivotal production node, it must possess the capacity for on-orbit manufacturing and production of space equipment with definable functions, in addition to on-orbit spacecraft identification, localization, capture, and maintenance capabilities. This trend is of great significance for the improvement of future spacecraft production efficiency, the extension of service life, and the enhancement of human space exploration capability in the future.

8. Conclusions

All major spacefaring nations have made the requisite technical preparations for the implementation and development of on-orbit assembly of space robots. For example, the United States, Canada, Germany, Japan, China, and other countries have conducted comprehensive research on space robot systems. This research has not only involved the design and manufacture of space robot bodies but has also included the verification of their performance in the space environment. Concurrently, the viability of employing space robotics for a range of tasks, including space fuel refueling, the dismantling and reassembling of replaceable modules, and the inspection and maintenance of spacecraft, has been further substantiated through the implementation of diverse on-orbit projects. The experiments and verification of the relevant projects have also resulted in the development and maturation of a series of related technologies, including space data communication technology, remote observation technology, and ground remote operation technology.

Nevertheless, there has been a paucity of efforts to design and construct large-scale space platforms, as the overwhelming majority of design and production is still based on the ground environment as an implicit reference. The ISS and CSST, the two most significant large-scale space facilities currently in orbit, are assembled and constructed by launching functional modules into space using ground-based carrier rockets. These modules are then docked with the ISS or CSST using space robotic arms or automated rendezvous and docking technology. The primary functions of these facilities are to conduct space experiments and maintain the lives of astronauts in orbit. In contrast, major spacefaring nations, such as the United States, Germany, and China, have proposed conceptual projects such as SpiderFab, Space Factory 4.0, and the autonomous construction of space telescopes. However, there is still a paucity of experience in designing and constructing large-scale autonomous space platforms using the space environment as a reference.

As artificial intelligence continues to evolve, it is likely that more advanced facilities and equipment will have a transformative impact on the future of space exploration and resource utilization. Facility intelligence will serve as the foundation for space activities, facilitating the construction and maintenance of an autonomous framework for space system operations. The intelligence of equipment will complement the intelligence of facilities, thereby further refining and expanding their operational capabilities. The organic integration and interaction of these two elements will serve as an important foundation for the future implementation of autonomous, unmanned space activities and resource utilization.