Leveraging Passive Compliance of Soft Robotics for Physical Human-Robot Collaborative Manipulation
Dallin L. Cordon, Shaden Moss, Marc Killpack, John L. Salmon · 2025
Read Leveraging Passive Compliance of Soft Robotics for Physical Human-Robot Collaborative Manipulation by Dallin L. Cordon, Shaden Moss, Marc Killpack, John L. Salmon (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-08184.txt.
Introduction
Dallin Cordon and three co-authors benchmark a large soft robot carrying a long object together with a human partner, in this 2025 paper. The robot is a three link pneumatic arm, bendy along its whole length, mounted on a mobile base that rolls in any direction. Softness promises safety around people, but strength was doubted.
The study mimics earlier experiments where two people carry an object, one leading with sight, one following blind. The robot played the follower role. Its performance, measured and judged, compared well with human followers in the same setup. The results push back on the assumption that soft robots are too weak for real carrying work. The authors see promise for jobs needing both strength and gentle adaptability, from warehouses to care settings.
Catalog record
| CREATOR | Dallin L. Cordon, Shaden Moss, Marc Killpack, John L. Salmon |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-08184
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Leveraging Passive Compliance of Soft Robotics for Physical Human-Robot Collaborative Manipulation Dallin L. Cordon, Shaden Moss, Marc Killpack, John L. Salmon Abstract . This work represents an initial benchmark of a large-scale soft robot performing physical, collaborative manipulation of a long, extended object with a human partner. The robot consists of a pneumatically-actuated, three-link continuum soft manipulator mounted to an omni-directional mobile base. The system level configuration of the robot and design of the collaborative manipulation (co-manipulation) study are presented. The initial results, both quantitative and qualitative, are directly compared to previous similar human-human co-manipulation studies. These initial results show promise in the ability for large-scale soft robots to perform comparably to human partners acting as non-visual followers in a co-manipulation task. Furthermore, these results challenge traditional soft robot strength limitations and indicate potential for applications requiring strength and adaptability. DOI: XXXXXXX.XXXXXXX Journal: THRI Volume: 37 4 111 2 CCS: Computer systems organization Robotic control CCS: Computer systems organization Robotic components CCS: Computer systems organization Robotic autonomy CCS: Hardware Physical verification CCS: Hardware Emerging technologies CCS: Human-centered computing Human computer interaction (HCI) CCS: Human-centered computing Collaborative and social computing Figure 1. Collaborative manipulation between a human and a large-scale soft robot. A human (left) collaboratively holding a stretcher-like object with a robot (right). Side view diagram showing a human (left) collaboratively holding a stretcher-like object with a robot (right). The robot consists of a soft arm with three continuum bellows joints mounted on a mobile base. 1. Introduction Collaborative manipulation of large, heavy, or unwieldy objects (defined in this paper as co-manipulation) has likely been a fundamental aspect of human interaction for millennia. As we seek to develop robotic helpers that operate more directly with humans, human-robot co-manipulation presents a unique challenge. For robots to work effectively with humans in shared object-carrying tasks, they must be designed to both complement human actions and align with human cognitive models so the robots’ interactions become intuitive and natural for their human partners. Furthermore, it is imperative that robots ensure the safety of their human counterparts by adjusting to unpredictable movements and disturbances while avoiding excessive force. The field of soft robotics has been presented as a highly viable solution to enable effective and safe physical human-robot interactions. The inherent compliance of soft robots is ideal for safe interactions as their flexible and adaptive structures can absorb impact and reduce the risk of injury during contact. This makes them particularly well-suited for tasks that require close physical proximity to humans, where safety and adaptability are paramount. However, much of the work in soft robotics has been concentrated on silicone-based robots, which tend to have low force output and limited power. Other work has focused on grippers designed to manipulate fragile objects of low mass and volume ( Shintake et al., 2018 ) , truss-based robots ( Usevitch et al., 2020 ) , expanding robots engineered to navigate cluttered environments ( Hawkes et al., 2017 ) , and wearable robots ( Zhu et al., 2022 ) . Few researchers have demonstrated soft robots with capabilities for external, heavy-duty collaborative manipulation tasks with humans. Figure 2. Physical human-soft-robot co-manipulation with a passively compliant robot arm. Photograph of a human (right) collaboratively holding a stretcher-like object with a robot (left). Photograph of a human (right) collaboratively holding a stretcher-like object with a robot (left) in a hallway. The robot consists of a soft arm with three continuum bellows joints mounted on a mobile base. This work demonstrates the potential of large-scale soft robotics for effective human-soft-robot interaction through a novel system: a passively compliant soft-robotic manipulator mounted on a custom omni-directional mobile base (Figure 2 ). We evaluate the system’s performance through human-soft-robot co-manipulation tasks involving a long, extended object and compare that performance to previous human-human co-manipulation experiments with similar payloads ( Shaw, 2024 ; Freeman et al., 2024 ) . In our study, nineteen individuals were recruited to perform planar translations (forward and back, side-to-side, and diagonal) of the co-manipulation object (CMO) while immersed in a virtual environment. The translational tasks were chosen to match a subset of the tasks performed by Shaw ( Shaw, 2024 ) to enable direct performance comparison. Participants completed sixteen sets of randomized translational tasks, each under different robotic parameters. The NASA Task Load Index (NASA-TLX) was administered after each set to assess perceived workload. While the full parameter study is being analyzed separately, this work focuses on four sets using consistent robot parameters from which we perform comparisons with previous human-human studies. Interaction between the robot and the human took place solely through the CMO, with no direct physical contact. The human operator possessed exclusive knowledge of the object’s desired trajectory and environmental constraints allowed for the existence of a direct, feasible path from the initial position to the desired goal. The robotic arm actively contributed to support some of the weight of the object throughout the co-manipulation task. Survey questions from Freeman et al. ( Freeman et al., 2024 ) were administered to participants both before and after the experiment. Shaw ( Shaw, 2024 ) employed a subset of Freeman’s post-study questions in their subsequent research, enabling qualitative analysis across all three studies. In short, the main contributions of this work are: • The development of a novel mobile soft-manipulation robotic platform for human-soft-robot collaborative manipulation. • A benchmark study for physical human-soft-robot co-manipulation with a soft robot arm. • An analysis of our robot partner’s co-manipulation performance relative to human partner performance for similar tasks. We seek to lay the foundation for future work focused on optimizing the control and coordination of these soft mobile manipulators for co-manipulation tasks in order to further improve user acceptance and team efficacy. The remainder of this paper is structured as follows: Section 2 provides background information and context for our work. Section 3 details our methodology, including the system’s hardware design - specifically the mobile base and robotic arm - along with the control architecture and the design of our experiment. Finally, Sections 4 and 5 present the experimental results and their associated discussion. 2. Background In co-manipulation tasks with rigid manipulators, compliance-based control is essential to address the complex interplay between the manipulator and the object. Challenges may arise from uncertainties in both the robot’s kinematic model and the grasping technique. These uncertainties can generate unexpected internal forces between the robots’ end-effector and the manipulated object, potentially compromising task performance and safety. Impedance control—and its counterpart, admittance control—as pioneered by Hogan ( Hogan, 1984 ) , has emerged as a prevalent approach in co-manipulation control strategies, enabling robots to effectively respond to positional deviations or interaction forces at the end-effector with the appropriate mechanical compliance ( Keemink et al., 2018 ; Song et al., 2019 ) . Ikeura ( Ikeura and Inooka, 1995 ) presented one of the first works in co-manipulation using a variable impedance controller. Since then, researchers have expanded on this concept to enhance co-manipulation performance. Duchaine et al. introduced an approach implementing force derivatives for human intention estimation and showed that velocity-based variable impedance control outperformed position-based approaches for human-soft-robot collaboration ( Duchaine and Gosselin, 2007 ) . Gopinathan et al. explored adapting stiffness