Reconfigurable joint limits and workspaces

Abstract

A modular, reconfigurable hardware workspace mechanism is provided, referred to as a Reconfigurable Joint Limit Module (RJLM), that physically alters the range of motion for each robot joint. By reconfiguring joint limits at the hardware level, the RJLM enables safer and more predictable robot operation in human-shared environments, independent of continuous software supervision.

Description

[0001] RECONFIGURABLE JOINT LIMITS AND WORKSPACES

[0002] FIELD OF THE INVENTION

[0003] This invention relates to methods, devices and systems for reconfigurable physically set joint limits and workspaces.

[0004] BACKGROUND OF THE INVENTION

[0005] A long-standing goal in robotic manipulation has been to enable robots to interact physically with humans in a way that is not only precise and compliant, but also inherently safe. As robots increasingly enter human-shared environments — public spaces, hospitals, homes — new constraints arise: they must minimize the risk of injury during and near contact.

[0006] While advancements in control and sensing technologies have improved robotic performance, safety concerns remain, particularly in tasks where a general-purpose industrial robot’s default workspace significantly exceeds the specific region of interest. Even minor errors can cause the robot to enter unintended areas, posing risks to nearby people or objects.

[0007] Conventional approaches rely heavily on software-defined virtual boundaries to constrain motion. While effective in principle, such methods critically depend on uninterrupted and error-free operation of the software and its controllers. Factors such as incorrect sensor readings, software bugs, electromagnetic interference, or sudden physical disturbances present significant challenges, potentially undermining safety mechanisms and leading to hazardous outcomes.

[0008] To address these limitations, the present invention introduces a modular, reconfigurable hardware workspace mechanism — referred to as a Reconfigurable Joint Limit Module (RJLM).

[0009] SUMMARY OF THE INVENTION

[0010] The present invention introduces a modular, reconfigurable hardware workspace mechanism - referred to as a Reconfigurable Joint Limit Module (RJLM) - that physically alters the range of motion for each robot joint. By reconfiguring joint limits at the hardware level, the RJLM enables safer and more predictable robot operation in human-shared environments, independent of continuous software supervision.

[0011] This invention is based on the concept of modifying the range of motion of a robot joint by reconfiguring its hardware joint limits. It introduces a Reconfigurable Joint Limit Module (RJLM), a mechanical device that enables this functionality by physically adjusting the positions of two structural limiters that define the joint’s motion range — angular constraints for revolute joints and linear bounds for prismatic joints.

[0012] The Reconfigurable Joint Limit Module (RJLM) is a compact, modular mechanism designed to mount directly onto a robot joint actuator. It physically enforces configurable upper and lower joint limits using a ratchet-based mechanical structure. The adjustment of these limits is performed electromechanically; however, once the desired range is set, the mechanism passively locks into place through purely structural means. During operation, the RJLM remains completely unpowered and electrically isolated from the robot's main control system. As such, it functions as a purely physical constraint, intrinsically limiting joint motion without requiring continuous software supervision or electrical input. Accordingly, this reconfiguration is intended to occur during a pre-task phase. Once the taskspecific workspace is defined — based on the spatial location of the task subject or target and the required end-effector trajectories — the corresponding joint range is computed. The RJLM then adjusts its two bounds to match this computed joint range and remains fixed throughout task execution.

[0013] By physically constraining joint motion to the task-relevant workspace, this approach enhances operational safety and predictability, while also promoting passive compliance.

[0014] In one embodiment, the present invention can be characterized as a method to mechanically adjust a workspace for a multi-degree of freedom robotic system relative to an object. In this method one would having a multi-degree of freedom robotic system with a joint for each of the multi-degree of freedoms. Each of the joints has a configurable joint limit module, where each of the configurable joint limit modules has at least one mechanical adjustable hard stop. A first computer integrated module is used for identifying a boundary of the object. A second computer integrated module is used for calculating joint limits for each of the joints, therewith calculating a workspace for the multi-degree of freedom robotic system, where this calculation uses dimensions obtained from the identified body boundary. Then the method mechanically sets the at least one mechanical adjustable hard stop for each of the configurable joint limit modules based on the calculated recommended joint limits. The multi-degree of freedom robotic system is then operated relative to the body using at least one mechanically adjustable set hard stop for each of the configurable joint limit modules.

[0015] Each of the configurable joint limit modules has a mechanical ratchet and a pawl locking mechanism, where the pawl locking mechanism includes a two-directional disengagement groove that enables a bidirectional adjustment under an active actuation state and a structural and passively locking when in an unpowered state.

[0016] Each of the configurable joint limit modules is structurally and electrically decoupled from a joint's main actuation system.

[0017] Each of the configurable joint limit modules is removably attached to its respective joint such that an original joint structure remains functionally unaltered when the respective configurable joint limit module is removed.

[0018] The joint limits are selected to avoid kinematic singularities and to preserve kinematic redundancy throughout task execution.

[0019] The workspace is configured to approximate a task-specific region while allowing a user-defined penetration tolerance into restricted zones.

[0020] In another embodiment, the present invention can be characterized as structure or device or system having the structural components and mechanisms as described herein.

[0021] Embodiments of the invention have at least the following advantages.

[0022] Intrinsic safety: Joint limits are physically enforced via mechanical components, independent of software, sensors, or power. This provides a fail-safe mechanism that is robust against control errors, sensor faults, and power failures.

[0023] Task-specific configurability: The reconfigurable design allows the joint limits to be adjusted for each task, enabling dynamic adaptation of the robot’s reachable workspace to enhance both safety and efficiency.

[0024] Passive locking: Once configured, the limiters mechanically lock in place without requiring continuous actuation or electrical power, reducing energy consumption and eliminating drift due to mechanical play or software error.

[0025] Modular architecture: The RJLM can be integrated into each joint of a multi -DOF robot without modifying the underlying actuator or control architecture, making it applicable to existing robotic platforms.

[0026] Redundancy-preserving workspace shaping: For redundant manipulators, such as 7-DOF arms, the module allows workspace reduction without collapsing the nullspace, thus retaining mani pulability and configuration flexibility within constrained joint ranges.

[0027] BRIEF DESCRIPTION OF THE DRAWINGS

[0028] FIG. 1 shows according to an exemplary embodiment the overall architecture of the Reconfigurable Joint Limit Module (RJLM), a revolute joint type, providing a high-level view of the complete assembled system. Detailed descriptions of individual subcomponents and mechanisms are provided in subsequent figures.

[0029] FIG. 2 shows according to an exemplary embodiment a revolute joint configuration with a revolute joint type RJLM installed, including the input section (i), the output section (ol), and the fixed hard stop element (h)

[0030] FIG. 3 shows according to an exemplary embodiment the revolute joint with separated input (i) and output (ol) sections. The hard stop (h) is fixed to the input side and moves within a range (rl) bounded by two adjustable mechanical limiters: the lower limit (11) and the upper limit (12). This figure illustrates how the joint range of motion is physically constrained within a reconfigurable window.

[0031] FIG. 4 shows according to an exemplary embodiment a standard hard-stop configuration in conventional robots, using elements h and h2. The input section (i) and output section (o2) are visible, with the hard stop (h) located on the input side and contacting the limiter (h2) on the output side.

[0032] FIG. 5 shows according to an exemplary embodiment a detailed breakdown of FIG. 4 by isolating the input (i) and output (o2) sides. The hard stop (h), mounted on the input section, moves within a constrained range (r2) defined by two fixed limiters (hl, h2) located on the output section.

[0033] FIG. 6 shows according to an exemplary embodiment an exploded view of the RJLM, showing both the Actuation Units (2a, 2b) and Anchoring Units (la, lb), along with the movable joint limiters (11, 12).

[0034] FIG. 7 shows according to an exemplary embodiment the internal ratchet mechanism within the anchoring unit. The ratchet gear (20) is constrained by asymmetric tooth faces (20a, 20b) and engaged by a pawl (21) mounted on the limiter plate (23), which also includes a secondary offset pawl tooth (23a) for finer resolution. The upper joint limiter (12) is attached to the same plate.

[0035] FIG. 8 shows according to an exemplary embodiment a pawl’s disengagement groove, which enables two-directional locking and adjustment. The peg (13) interacts with the inclined face (21a) and horizontal face (21b) of the pawl (21), which is supported by two springs (22a, 22b), and engages with the ratchet gear (20) through tooth faces (20a, 20b).

[0036] FIG. 9 shows according to an exemplary embodiment the pawl component (21) in isolation, illustrating its internal disengagement groove, including the inclined face (21a) and horizontal face (21b).

[0037] FIG. 10 shows according to an exemplary embodiment the actuation unit, which includes the motor (ml), encoder (el), worm screw (10), worm gear (11), peg-driving wheel (12), and the peg (13) that interfaces with the limiter mechanism.

[0038] FIG. 11 shows according to an exemplary embodiment a conventional prismatic joint without joint limit modification. The joint comprises a translating actuator (11) and a housing (It).

[0039] FIG. 12 shows according to an exemplary embodiment the prismatic j oint at its lower limit, with the actuator component (11) fully retracted into the housing (It).

[0040] FIG. 13 shows according to an exemplary embodiment a prismatic joint equipped with an RJLM , featuring an adjustable upper limit using a limiter component (12) mounted on the actuator (11).

[0041] FIG. 14 shows according to an exemplary embodiment a 7-DOF robotic arm (31) performing an ultrasound scan on a human subject (30), using an ultrasound probe (31p). The robot is positioned relative to its surrounding environment (32).

[0042] FIG. 15 shows according to an exemplary embodiment the unconstrained full workspace (34a) of the robot without any joint limit configuration, encompassing both the target region and surrounding space. FIG. 16 shows according to an exemplary embodiment the reshaped and constrained workspace (34b) resulting from task-specific joint limit settings using the RJLM

[0043] DETAILED DESCRIPTION

[0044] The core of this invention lies in the ability to mechanically adjust and reduce the joint movement range in multi-degree-of-freedom robotic systems. The developed Reconfigurable Joint Limit Module (RJLM) is a compact, modular device designed for integration into robotic joints (FIG.

[0045] 1 and FIG. 11). It enables mechanical adjustment of both the upper and lower bounds of a joint’s motion range through an embedded mechanism.

[0046] Although the adjustment is performed electromechanically, once the joint limits are set to the designated values, the mechanism can be passively locked in place using only structural components, without requiring electrical power. This structural -level locking ensures reliable enforcement of the configured limits during task execution.

[0047] Robotic joints are generally classified as either revolute or prismatic. Revolute joints produce rotational motion about a fixed axis, whereas prismatic joints enable linear translation along a single direction. Due to the fundamental mechanical differences between these two types, the RJLM has been developed in two distinct variants — each specifically tailored to either revolute or prismatic joints. While they share the same conceptual principle and overall functional goals, their mechanical implementations differ significantly.

[0048] Accordingly, the following sections describe the designs and mechanisms of each RJLM type separately. Reconfigurable Joint Limit Module (RJLM) on Revolute joint

[0049] Ideally, revolute joints can rotate beyond 360 degrees and encompass the entire scope. However, excessive rotation in one direction can lead to electrical wire twisting and potential damage or disconnection, as well as increase the risk of self-collision among the robot’s body parts.

[0050] To mitigate these risks, conventional robots often employ simple mechanical stops (h, h2) the joints' inputs (i) and outputs (o2) to limit the range of motion as illustrated in FIGs 4-5.

[0051] In this context, the input (i) refers to the driving section of the robot joint, which contains the electrical motor responsible for the joint's movement. The frame of the input section is fixed to the base or the preceding joint. The output section (o2) refers to the part that is moved by the motor within the input section (i). This output can be connected to the input of the subsequent joint and is also referred to as the robot link.

[0052] These hard stop components, or metal protrusions, (h, h2) are secured to the joints using screws, serving as static joint range limiters. The hard stop (h) on the input structure (i) moves along with the rotation of the joint. When it contacts the hard stop (h2) on the output structure (o2), it restricts further rotation, thereby limiting the joint range to r2. The range r2 can be adjusted by altering the size and position of h2.

[0053] The primary distinction of a revolute joint with the RJLM is the use of 11 and 12 instead of h2, as shown in FIGs. 2-3. Here, 11 is a metal protrusion that determines the position of the lower joint limit, and 12 is a metal protrusion that determines the position of the upper joint limit. As the robot joint operates, h travels only between 11 and 12, causing the joint range (rl) to vary according to the positions of 11 and 12. Therefore, the primary function of the RJLM is to adjust and securely lock the positions of 11 and 12.

[0054] The following detailed description of the mechanism is provided with reference to the drawings. As shown in FIG. 6, this mechanism can be divided into two main sections: the Actuation Units (2a, 2b), which move the two limiter wings (11, 12), and the Anchoring Units (la, lb), which secure the wings in place after movement. Each limit requires a separate section, with la as the Locking Unit and 2a as the Actuation Unit for 11, and similarly, lb and 2b for 12. Together, these two pairs constitute a single RJLM.

[0055] The critical aspect of this mechanism is that 11 and 12 must be able to move and then remain fixed in place without electrical power. To achieve this, the Anchoring Units (la, lb) employ a specialized ratchet mechanism distinct from conventional ratchet systems, particularly in the design and functionality of the pawl (21).

[0056] FIG. 7 provides a detailed view of lb, one of the actuation units. The pawl (21) is connected to and supported by two springs (22a, 22b) and can move radially — or vertically in FIG. 7 — to engage or disengage with the ratchet gear (20). Thus, unless specifically controlled otherwise, it functions like a typical ratchet mechanism. The ratchet gear is free to move in the direction of the inclined face of the teeth (20a) but is restricted when moving in the direction of the vertical face (20b). By altering the position of the pawl (21), the joint limiters (11, 12) are locked in place.

[0057] However, the joint limiters should be able to adjust their position to any value. To achieve this, a typical ratchet mechanism, which allows movement in only one direction (20b), cannot be used for the purposes of the RJLM.

[0058] To address this, the design of the pawl (21) was modified as shown in FIGs. 8-9, enabling 12 to be moved in 20b direction as well. This specialized pawl (21) has a uniquely shaped groove on one side. By inserting a small peg (13) into this groove and moving it horizontally, the interaction between the peg and groove can disengage the pawl (21).

[0059] This groove is shaped with an incline (Ila) towards the direction (20b) and horizontally (21b) towards the direction (20a). Imagine a peg centered on the pawl’s groove shown in FIG. 8 and rotates concentric to the limiter plate (23). When the peg contacts 21a, it pushes down the incline (21a), disengaging the pawl (21). If the peg continues to move in that direction (20b), the pawl pushes and therefore rotates the limiter plate (23). This alters the pawl's position with respect to the ratchet gear (20) and thus reconfigure the upper joint limit, 12’ s orientation. Conversely, if the peg moves towards the horizontal groove (21b), it directly contacts the limiter plate (23), allowing 12 to rotate clockwise (20a) and changing the pawl's position.

[0060] In summary, with this innovative pawl design, the limiter plate (23) can only rotate in one direction (20b) without using the peg (13). However, using the peg (13) with the pawl's groove (21) allows the limiter plate (23) to rotate in both 20a and 20b directions with respect to the ratchet gear (20). This feature ensures that once the angle limit value is set and the limiter is engaged, it can maintain its position without the need for electrical power, securing the set limit indefinitely until reconfigured. The mechanism of lb is also present and applicable in la.

[0061] As shown in FIG. 7, a single limiter plate includes two pawl mechanisms (21), which are not symmetrically placed. Due to the use of a ratchet mechanism, the RJLM's resolution is determined by the size of the ratchet wheel's teeth, which can make very fine angle adjustments challenging. By offsetting the pawls by half the size of the teeth (23a), the resolution is effectively doubled, allowing for finer adjustments.

[0062] Shown in FIG. 10, the actuation unit (2a, 2b) is responsible for electrically moving the peg that manipulates the pawl in the anchoring unit, described in the previous section. The actuation unit comprises a motor (ml), an encoder (el), a worm screw (10), a worm gear (11), a peg (13), and a driving wheel (12) that holds the peg. The peg (13) is attached to the driving wheel, which is directly connected to the worm gear. The worm gear is driven by the worm screw, which is directly coupled to the motor at a 1 : 1 ratio.

[0063] By controlling the motor, the peg is rotated, thereby adjusting the position of the pawl (21) and ultimately setting the value of the joint limit, orientation of 11 or 12.

[0064] 2a and 2b have identical structures and mechanisms.

[0065] When the robot joint attempts to move beyond the limited range, the ratchet mechanism of the anchoring unit effectively restricts the joint motion. Yet, the use of a worm gear with self-locking capability in the actuation unit enhances safety. Furthermore, the minimal backlash of the worm gears contributes to the stability of the system. Reconfigurable Joint Limit Module (RJLM) on Prismatic Joint

[0066] The linear joint actuator mechanism of a prismatic joint is typically structured as shown in FIGs. 11-13. Accordingly, as illustrated in FIG. 11, the lower limit of a conventional prismatic joint is achieved when the actuated component (11) fully retracted into the housing (It) component. Conversely, the upper joint limit is reached when 11 is fully extended. Since the position where 11 is completely retracted into It results in minimal interference with humans, there is no necessity to raise the lower limit of the prismatic joint. Therefore, only the method for reducing and reconfiguring the upper joint limit is described.

[0067] As depicted in FIG. 11, 12 is mounted on 11. The outer diameter of 12 is larger than the opening of It. Thus, by adjusting the position of 12, which is coupled to 11, the upper joint limit can be controlled.

[0068] The linear joint actuator mechanism comprising the prismatic joint generally appears as depicted in FIGs. 11-13. Thus, as shown in FIG. 11, the actuated component (11) fully retracted into the housing (It) component represents the lower limit of a typical prismatic joint. Conversely, when the actuated component (11) is fully extended, it represents the upper joint limit. Since the state where 11 is fully retracted into It is the least interfering position for the robot with respect to humans, there is no need to elevate the lower limit of the prismatic joint. Therefore, only the method to reduce and reconfigure the upper joint limit is described.

[0069] In FIG. 11, 12 is mounted on 11. The outer diameter of 12 is larger than the opening of It. Thus, by altering the position of 12, which is coupled to 11, the upper joint limit can be controlled. Reconfigurable Workspace

[0070] A reconfigurable workspace refers to the concept that the task space of the robot’s end-effector can be reshaped by adjusting joint limits using Reconfigurable Joint Limit Modules (RJLMs). Rather than being fixed, the shape and extent of the workspace can be reconfigured according to the specific requirements of a task or environment, allowing unnecessary and excessive portions of the workspace to be eliminated for safety.

[0071] This concept is grounded in the idea that by mechanically restricting the joint limits, the task space of the robot’s end-effector can be physically constrained to a range that ensures safety while still preserving the motion capabilities required to complete the task.

[0072] This functionality becomes especially critical in applications involving human-robot interaction or shared environments. Since the range of motion directly affects operational safety, configuring joint limits appropriately is essential. If the joint limits are set too wide, the robot may enter undesired areas; if set too narrowly, the robot's manipulability and task coverage may be compromised. Therefore, selecting joint limit ranges that correspond to the user’s desired safety level is key to achieving safe and efficient operation.

[0073] To reconfigure the robot’s workspace for a specified task, the task-workspace is initially defined as the area within which the end-effector must operate, while concurrently identifying prohibited regions, areas that must be avoided due to safety constraints. The objective is to design the robot’s reachable workspace to completely encompass the task region, while preventing entry into prohibited zones. However, since RJLMs impose constraints on joint motion through fixed mechanical limits at each joint, the resultant reachable workspace cannot be precisely tailored to fit arbitrary taskworkspace boundaries, particularly those that are flat or convex. To address this limitation, a taskspecific penetration tolerance is established, and the joint limit configuration is optimized to better approximate the desired workspace within this acceptable margin.

[0074] For example, consider a typical 7 DOF robot arm (31) with all revolute joints, equipped with an ultrasound scope (31p) at its end-effector to scan a patient's abdomen (30), as shown in FIG. 14. The task workspace includes the region in direct contact with the abdomen, permitting for a few centimeters of penetration. Prohibited spaces include the body volume or human occupied space. Without an RJLM, the robot's workspace (34a), as depicted in FIG. 15, would cover not only the patient's body but also a large radius around the robot. Consequently, any incorrect movement by the robot could severely injure not only the patient being scanned but also anyone nearby.

[0075] Limiting the workspace using an RJLM, as illustrated in FIG. 16, significantly reduces the workspace (34b) and increases overall safety. Specifically, this workspace is designed so that the robot's end-effector cannot penetrate downward towards the body, thus enhancing the safety of the person being scanned.

[0076] The method is as follows: First, the boundaries of the person or patient to be scanned with an ultrasound are identified using a camera or any device capable of 3D spatial recognition. Depending on which organ of the person is to be scanned, a pre-trained program calculates the necessary position and orientation of the robot's end-effector based on the approximate scan location, depth, and angle required for the probe. Then, the workspace of the robot is determined to ensure it can perform the scan without exceeding necessary limits, and the joint limits are calculated accordingly. Each joint limit is then adjusted to reflect these calculations. When the robot is operated, it can be used to scan the person safely within an optimized workspace.

[0077] Given that the robot used for this task is a 7 DOF robot with redundancy, there are numerous ways to set the joint limits for the same task. The fundamental approach, however, is to set the joint limits so that the robot's movements do not approach a singularity configuration. By doing so, the robot can maintain various configurations, allowing for more freedom of movement within the restricted joint range while effectively limiting the workspace. Additionally, it's preferable to use a larger range of the front joints, as they have less effective inertia, to ensure the robot does not exert excessive force on the body during scanning. Additionally, it's important to position the robot's base to avoid singularity configurations and to ensure that the robot's default configuration is reconfigured to align well with the scan area and workspace.

[0078] Based on these principles, several workspace shapes and joint limit values are presented to the operator, allowing them to choose and set the degree of movement restriction based on the patient's condition and the required scan depth and width. This ensures an appropriate level of safety for each scan. Moreover, when the robot base position is adjustable relative to the subject, a broader set of joint limit combinations becomes feasible for workspace adaptation.

Claims

CLAIMSWhat is claimed is:

1. A method to mechanically adjust a workspace for a multi-degree of freedom robotic system relative to an object, comprising:(a) having a multi-degree of freedom robotic system with a joint for each of the multidegree of freedoms, wherein each of the joints comprises a configurable joint limit module, wherein each of the configurable joint limit modules comprises at least one mechanical adjustable hard stop;(b) a first computer integrated module for identifying a boundary of the object;(c) a second computer integrated module for calculating j oint limits for each of the j oints, therewith calculating a workspace for the multi-degree of freedom robotic system, wherein the calculating uses dimensions obtained from the identified body boundary;(d) mechanically setting the at least one mechanical adjustable hard stop for each of the configurable joint limit modules based on the calculated recommended joint limits; and(e) operating the multi-degree of freedom robotic system relative to the body using at least one mechanically adjustable set hard stop for each of the configurable joint limit modules.

2. The method as set forth in claim 1, wherein each of the configurable joint limit modules comprises a mechanical ratchet and a pawl locking mechanism, wherein the pawl locking mechanism includes a two-directional disengagement groove that enables a bidirectional adjustment under an active actuation state and a structural and passively locking when in an unpowered state.

3. The method as set forth in claim 1, wherein each of the configurable joint limit modules is structurally and electrically decoupled from a joint's main actuation system.

4. The method as set forth in claim 1, wherein each of the configurable joint limit modules is removably attached to its respective joint such that an original joint structure remains functionally unaltered when the respective configurable joint limit module is removed.

5. The method as set forth in claim 1, wherein the joint limits are selected to avoid kinematic singularities and to preserve kinematic redundancy throughout task execution.

6. The method as set forth in claim 1, wherein the workspace is configured to approximate a task-specific region while allowing a user-defined penetration tolerance into restricted zones.

Images