The pursuit of sustained human and robotic presence in space necessitates the development of advanced on-orbit servicing (OOS) capabilities. Space robots, particularly those equipped with dexterous manipulator arms, are pivotal tools for these missions, performing tasks ranging from assembly and refueling to maintenance and debris capture. A critical observation from existing systems like the Space Station Remote Manipulator System (SSRMS) or the European Robotic Arm (ERA) is their functional specialization. These systems are typically designed with a single, dedicated end-effector for a narrow set of cooperative tasks. However, the future of OOS demands versatility to handle a broader spectrum of targets, including non-cooperative satellites or irregular debris. This article, from a designer’s perspective, presents a conceptual framework for a multifunctional OOS robot system. Its core innovation lies in a tool-changing architecture enabled by a primary end-effector, which in turn can deploy specialized secondary end-effectors tailored for both cooperative and non-cooperative capture scenarios.
The proposed robotic servicing system is centered around a free-flying spacecraft serving as a mobile base. Attached to this base is a 7-degree-of-freedom (7-DOF) redundant manipulator, chosen for its enhanced dexterity and ability to avoid singularities in complex workspaces. The manipulator’s symmetric design, with identical end-effectors at both the “wrist” and “shoulder,” enables a unique “inchworm” locomotion capability. This allows the entire arm to traverse the service spacecraft by sequentially grasping and releasing standardized interfaces, thereby greatly extending its effective operational envelope without propulsive movement of the base spacecraft.
The linchpin of this multifunctionality is the primary end-effector. This device must serve three primary functions: (1) as a robust mechanical interface for the inchworm gait, (2) as a tool-changer for swapping mission-specific end-effectors, and (3) as a conduit for power, data, and mechanical drive to the attached tools. The design that meets these requirements is a Three-Finger-Three-Petal (TFTP) end-effector. Its capture mechanism is based on three fingers, each guided by a unique cam profile machined into its structure. A single rotary actuator drives a lead screw, which translates a central platform. This translation, constrained by guide pins riding in the finger cams, orchestrates a precise sequence of motions: finger extension, petal splaying for misalignment accommodation, closure for capture, and final retraction to achieve a rigid, preloaded connection with the target interface.
The mathematical model governing the finger motion can be derived from the cam profile. Let the cam curve for one finger be defined in polar coordinates relative to the guide pin as $$ r(\theta) $$, where $$ \theta $$ is the rotation angle of the finger assembly (though actuated linearly, the motion is kinematically converted). The linear displacement of the drive platform, $$ s $$, is directly proportional to the actuator rotation. The relationship between $$ s $$ and the finger’s opening diameter $$ D_{grasp} $$ is critical for modeling the capture envelope. A simplified geometric analysis for the splayed position yields:
$$ D_{grasp}(s) = 2 \cdot (L_{finger} \cdot \sin(\alpha(s)) + r_{offset}) $$
where $$ L_{finger} $$ is the effective finger length, $$ \alpha(s) $$ is the splay angle determined by the cam slope at displacement $$ s $$, and $$ r_{offset} $$ is a fixed radial offset. This equation highlights how the design generates a large capture envelope (high $$ D_{grasp} $$) during the initial phase, accommodating significant positional and angular misalignments—a property known as high “capture tolerance.”
Furthermore, the TFTP end-effector integrates a concentric drive shaft within its central lead screw. This “socket wrench” can transmit torque to a tool or interface, enabling active actuation of a grasped payload or tool. Electrical and data connections are established via concentric ring connectors upon final docking. The key specifications of this primary end-effector are summarized below:
| Parameter | TFTP End-Effector Specification |
|---|---|
| Primary Function | Tool Change, Inchworm Locomotion, Utility Conduit |
| Capture Mechanism | Cam-Guided Three-Finger Latching |
| Actuation | Single Motor via Lead Screw |
| Key Feature | Integrated Power/Data/Mechanical Drive Transmission |
| Target | Standardized Cooperative Interface (TFTP-Compatible) |
The true versatility of the system is unlocked when the TFTP end-effector is used to pick up specialized secondary end-effectors from storage docks on the service spacecraft. Two such specialized end-effectors are proposed for distinct capture missions.
The first is a Steel Cable-Snared End-Effector (SCSEE), designed for the soft capture of objects with protruding elements, like a satellite’s engine nozzle (non-cooperative) or a dedicated grapple fixture (cooperative). Its operation is elegantly simple. Three cables are attached at one end to a fixed base and at the other to a rotating crown. A motor drives the crown, causing the cables to wind symmetrically inward, forming a tightening noose around the target. This winding action provides a compliant, force-limiting capture, drastically reducing impact loads on fragile targets. After capture, a second set of actuators (lead screws) engages to rigidize the connection by drawing the entire captured interface into a docking nest.
The capture tolerance here is defined by the initial loop diameter versus the target diameter. The major advantage is that the flexible cables can conform to irregular shapes and accommodate large lateral misalignments. The tensile force on the cable during soft capture, $$ T_{cable} $$, relates to the motor torque $$ \tau_{motor} $$ and the effective winding radius $$ r_{wind} $$:
$$ T_{cable} = \frac{\tau_{motor} \cdot G}{3 \cdot r_{wind}} $$
where $$ G $$ is the gear reduction ratio. The factor of 3 arises from the three cables sharing the load. This design ensures that the constriction force can be precisely controlled by the motor current, enabling a truly soft capture.
The second specialized tool is an Underactuated Three-Finger Adaptive Gripper (UTFG). This end-effector is intended for grasping non-cooperative, irregularly shaped objects such as orbital debris. Its fingers use a linked, differential mechanism. A single actuator closes all fingers simultaneously, but the kinematic linkage allows each finger to independently conform to the object’s shape until all fingers make contact and the grip is stabilized. This passive adaptation requires no sensory feedback for the conforming phase, making it robust and simple. The grasp force is self-regulated by the mechanism’s geometry and the applied actuator torque.
The kinematics of one underactuated finger can be modeled as a series of linkages. The force at the fingertip, $$ F_{tip} $$, is related to the input torque from the tendon or linkage, $$ \tau_{in} $$, through a Jacobian matrix $$ J $$ that depends on the instantaneous configuration of the finger joints $$ \vec{\theta} $$:
$$ \vec{F}_{tip} = (J(\vec{\theta})^T)^{-1} \cdot \vec{\tau}_{in} $$
This underactuation allows a complex, shape-conforming grip to be achieved with minimal actuation, saving weight and complexity—a critical advantage for space systems.
The operational workflow is as follows: The service spacecraft maneuvers to the target area. The 7-DOF arm uses its TFTP end-effector to perform inchworm locomotion to an optimal work site. For a cooperative target with a standard interface, the arm might dock directly using its TFTP end-effector. For a satellite nozzle capture, the arm first retrieves the SCSEE from its dock using the TFTP end-effector, then employs the SCSEE to perform the soft capture. For irregular debris, the UTFG is retrieved and deployed. Upon task completion, the specialized end-effector is returned to its dock, and the arm resumes its primary TFTP configuration or relocates.
The following table provides a comparative summary of the three proposed end-effector systems, highlighting their distinct roles and characteristics:
| Feature | TFTP End-Effector (Primary) | Steel Cable-Snared (SCSEE) | Underactuated Gripper (UTFG) |
|---|---|---|---|
| Primary Mission | Tool Change & Arm Locomotion | Soft Capture of Protrusions | Grasping Irregular Objects |
| Target Type | Cooperative Standard Interface | Cooperative/Non-Cooperative (Nozzle, Fixture) | Non-Cooperative (Debris) |
| Key Strength | Multifunctionality & Utility Transfer | High Tolerance & Force-Limited Soft Capture | Passive Shape Adaptation |
| Actuation Principle | Cam-Driven Finger Closure | Cable Winding & Lead Screw Retraction | Differential Linkage Mechanism |
| Compliance | Mechanical via Cam Profile | High (Flexible Cables) | Mechanical via Underactuation |
The performance of the end-effector, particularly its capture tolerance, can be quantified. For a given misalignment between the end-effector and the target interface, successful capture requires the target to lie within the “Capture Envelope Volume” (CEV). For the TFTP design, this is a complex volume dependent on the finger splay angle. A simplified metric is the Lateral Capture Tolerance $$ \delta_{lat} $$ at a given distance $$ z $$ from the end-effector face. For the initial splayed configuration, it can be approximated as:
$$ \delta_{lat}(z) \approx \frac{D_{grasp}(z) – D_{target}}{2} $$
where $$ D_{target} $$ is the diameter of the target interface. Maximizing $$ D_{grasp} $$ through mechanical design directly maximizes the system’s tolerance to positioning errors from the manipulator arm.
The development and testing of such systems require rigorous ground prototypes. Engineering models of the key end-effectors must be built to validate the kinematic models, actuation schemes, and capture sequences. The TFTP end-effector prototype would focus on the precision of the cam-guided motion and the reliability of the utility connections. The SCSEE prototype would be tested to measure the soft-capture force profile and the maximum tolerated lateral and angular misalignment before the cable fails to engage the target. The UTFG prototype would be evaluated on its ability to securely grasp a variety of geometric shapes without reconfiguration. Ground testing typically involves air-bearing tables to simulate two-dimensional microgravity dynamics and suspension systems for three-dimensional tests, allowing for the validation of the capture dynamics and the control algorithms that govern the approach and contact phases.
In conclusion, the future of on-orbit servicing lies in robotic systems that are adaptable, multifunctional, and capable of handling both planned cooperative tasks and unplanned, non-cooperative scenarios. The proposed architecture, centered on a tool-changing primary end-effector like the TFTP design, provides a viable pathway toward this goal. By enabling the on-demand deployment of mission-optimized tools such as the soft-capture SCSEE and the adaptive UTFG, a single servicing spacecraft can dramatically expand its portfolio of possible missions. This approach moves beyond the paradigm of single-purpose robots, offering a flexible and economically attractive solution for sustaining and expanding space infrastructure. The continued design refinement, prototyping, and validation of these advanced end-effector systems are essential steps in making versatile, multi-role on-orbit servicing robots a standard reality in the space operations ecosystem.