Source-linked AI summary
Design of a Biomimetic Joint-Covering Skin with Tissue-Like Structure to Enhance Proprioception in a Musculoskeletal Humanoid
Akihiro Miki, Shun Hasegawa, Yoshimoto Ribayashi, Kento Kawaharazuka, Kei Okada
TL;DR
Musculoskeletal humanoids commonly estimate proprioception from limited muscle sensing, leaving the contribution of cutaneous deformation around joints insufficiently explored. The study develops tissue-structured joint-covering skin with pressure- and stretch-sensitive elements, implements it on Musashi-W, and evaluates standalone and multimodal estimation. The skin alone estimates joint angle with an average error of approximately three degrees, while integration with muscle sensing improves accuracy and may help interpret external stimuli.
Problem
Proprioception in musculoskeletal humanoids often relies on limited muscle sensing, while the contribution of mechanoreceptors in joint-covering skin remains insufficiently explored.
Method
The study implements layered biomimetic joint-covering skin with receptor-like pressure- and stretch-sensitive elements on Musashi-W and evaluates it alone and with muscle sensing.
Results
The skin alone achieves joint-angle estimation with an average error of approximately three degrees, and fusion with muscle sensing achieves higher estimation accuracy than muscle-length-only sensing.
Takeaways & Limitations
Tissue-structured skin can extend proprioceptive systems by providing meaningful skin-derived joint information and complementary signals for interpreting external mechanical stimuli.
Takeaways & Limitations
The implementation uses fewer and substantially larger receptor-like sensors than biological mechanoreceptors, making it a conceptual rather than strict biological reconstruction.
Abstract
from arXiv · showhide
Proprioception in musculoskeletal humanoids is typically estimated primarily from muscle sensing, while the role of cutaneous deformation around joints remains insufficiently explored. In biological systems, mechanoreceptors distributed within soft tissue complement muscle feedback and support reliable joint state estimation. This study presents the design of a biomimetic joint-covering skin with a tissue-like layered structure that integrates pressure- and stretch-sensitive elements within the joint-covering tissue. The proposed skin is implemented on the musculoskeletal humanoid Musashi-W, and its independent proprioceptive capability as well as its integration with muscle sensing are evaluated. Experimental results show that the proposed skin alone achieves joint angle estimation with an average error of approximately 3 degrees. Furthermore, integration with muscle sensing improves estimation accuracy. Owing to its joint-covering structure, the skin may mechanically mitigate the influence of external disturbances on the muscles, and the integration of multiple modalities suggests the possibility of contributing to the identification of external stimuli that are difficult to interpret using muscle sensing alone. This work presents a design methodology for biomimetic joint-covering skin and demonstrates that such tissue-structured skin can serve as an effective approach for extending proprioceptive systems in musculoskeletal humanoids.
I. INTRODUCTION
Musculoskeletal humanoids need proprioception for stable, adaptive motion, yet joint-angle estimation often relies on limited muscle sensors. This study addresses that gap with tissue-structured joint-covering skin that combines skin-derived sensing with muscle sensing.
- Accurate proprioception is essential for stable behavior and adaptive motion in musculoskeletal humanoids.
- Joint-angle estimation is difficult in musculoskeletal humanoids because their complex structures prevent straightforward direct measurement.
- Existing approaches often rely on limited sensor numbers and modalities, including muscle-length and tension sensors.
- Biological proprioception integrates signals from mechanoreceptors in muscles, skin, and surrounding joints, providing redundancy and supporting discrimination of external disturbances.
- The study designs and evaluates layered joint-covering skin with pressure- and stretch-sensitive elements on Musashi-W, both independently and alongside muscle sensing.
II. RELATED WORKS
Prior tactile and embedded-sensing systems mainly target contact, tactile mapping, or task-specific shape reconstruction. This work instead adopts biological tissue organization and numerous receptor-like elements as a proprioceptive design principle.
- Biological skin distributes multiple mechanoreceptor types across layered tissue to integrate diverse tactile and proprioceptive information.
- Existing biomimetic tactile sensors primarily target object recognition and contact-state estimation rather than proprioception.
- Whole-body modular skins primarily support contact detection and tactile mapping, while proprioception is not their main focus.
- Flexible e-skins provide large-scale sensing but face implementation complexity and possible effects on joint mechanics when adapted to three-dimensional joints.
- Embedded sensors in soft bodies are typically arranged to achieve specific functions with a minimal number of informative sensors.
- The proposed design follows biological organization by embedding numerous receptor-like elements within supporting tissue rather than optimizing a grid-like sensor layout.
III. DEVELOPMENT OF THE TISSUE-STRUCTURED BIOMIMETIC JOINT-COVERING SKIN
The proposed skin uses a layered, tissue-level architecture rather than a small set of optimally placed high-performance sensors. Its materials and construction approximate increasing compliance from the outer to inner layers.
- The design embeds numerous receptors within a layered tissue structure that mechanically supports them.
- Fabrication embeds sensors during multilayer construction, combines paired skin components, coats the epidermis, encloses the joint, and mounts the structure on Musashi-W.
- The skin mimics epidermis, dermis, and subcutaneous tissue using a multilayer soft structure.
- Material selection targets the biological tendency for compliance to increase from the outer layer toward inner layers.
- The implementation uses silicone rubber for the epidermis, silicone gel for the dermis, and foamed silicone sponge for subcutaneous tissue.
B. Receptor-like Elements
The skin incorporates receptor-like elements selected for pressure- and stretch-related sensing and places them at different depths to reflect biological receptor organization.
- The study selects Ruffini-like stretch sensors and Merkel cell-like elements because spatial constraints prevent reproducing all biological cutaneous receptors.
- Strain gauges serve as Merkel cell-like elements by changing electrical resistance under applied strain.
- A conductive filament chainmail structure provides a stretch-responsive sensor corresponding to Ruffini endings.
- Merkel-like strain gauges are embedded between epidermal and dermal layers, while Ruffini-like structures are placed within the subcutaneous layer.
C. Development of Tissue-Structured Biomimetic Skin and Its Implementation in a Musculoskeletal Humanoid
The biomimetic skin reproduces layered biological tissue by embedding Merkel-like strain sensors, Ruffini-like conductive sensors, and fiber structures in a joint-covering soft tissue. It was fabricated as paired components and mounted over an open-structure biomimetic joint on Musashi-W.
- Fiber bundles embedded in the subcutaneous layer mimic collagen structures and provide mechanical continuity with the humanoid’s skeletal structure.
- Thirty-two strain-gauge sensors were positioned at the epidermis-dermis interface in a semi-random distribution modeled on biological receptor organization.
- Twelve conductive chainmail structures representing Ruffini-like elements were embedded in the subcutaneous layer and oriented along the fiber bundles.
- Dragon Skin 30 was coated and cured twice to form the epidermal layers, producing a pair of biomimetic skin components.
- The skin components enveloped an open-structure biomimetic joint and were connected at their interface using Soma Foama 15.
- Musashi-W integrates joint-covering skin signals with muscle-derived length and tension measurements as proprioceptive information.
IV. EXPERIMENTS USING TISSUE-STRUCTURED SKIN AND MUSCLE SENSING
Experiments collected synchronized muscle, skin, and motion-capture signals from Musashi-W during varied right-upper-limb postures. The resulting dataset supported elbow proprioceptive estimation across three degrees of freedom.
- Sensorimotor data were collected while operating Musashi-W equipped with the tissue-structured biomimetic skin.
- The right upper limb contains five muscle-module pairs actuating the shoulder and elbow, providing 10 muscle-length and 10 muscle-tension sensors.
- The skin contributed 32 Merkel-cell-like strain-gauge signals and 12 Ruffini-like stretch-sensor signals for estimating elbow joint angles.
- Motion capture replaced embedded potentiometers and measured elbow pitch, yaw, and roll angles using markers on the upper arm and forearm.
- Body-schema-driven motion produced varied postures despite the modified body structure, while the tissue-biomimetic joint enabled elbow roll motion.
- Approximately one hour of operation produced 615 sensorimotor samples covering about 50° of elbow motion along each axis.
- Muscle-related, cutaneous, and motion-capture measurements were recorded simultaneously to construct the proprioceptive-estimation dataset.
B. Single-Modality Proprioceptive Estimation
Single-modality experiments compared muscle and skin signals for three-degree-of-freedom elbow angle estimation using standardized learning and paired statistical evaluation. Muscle length performed best, while skin modalities achieved errors around three degrees.
- Three-degree-of-freedom elbow angles were estimated separately from muscle length, muscle tension, Ruffini-like skin, and Merkel-like skin signals.
- The dataset was split into 70% training, 15% validation, and 15% test sets across 20 random seeds.
- A fully connected MLP with three 256-unit hidden layers predicted the three joint angles using ReLU activations, dropout, and a linear output layer.
- RMSE was computed for each joint-angle dimension and averaged across dimensions, with distributions shown by modality and summary statistics reported across seeds.
- The Wilcoxon signed-rank test with Holm correction compared paired RMSE values across modalities because normality could not be reliably assumed.
- Muscle length had the smallest mean and median RMSE, significantly outperforming the other modalities with Holm-corrected p < 0.001.
- Approximately 3°: skin modalities generally produced RMSE values within a range of about three degrees, though larger than muscle length.
C. Multimodal Proprioceptive Estimation
Multimodal experiments compared muscle-only, skin-only, and combined inputs using concatenation and encoder-based fusion. Encoder-based fusion of both modalities significantly surpassed muscle-length-only estimation, unlike the other fusion conditions.
- The experiments defined muscle-only, skin-only, and combined muscle-plus-skin modality conditions.
- Concatenation fusion directly joined modality features before a shared MLP, whereas encoder fusion projected each modality into a 32-dimensional embedding before regression.
- Multimodal performance used RMSE from 20 random seeds and compared each condition against muscle length as the baseline with paired statistical testing.
- M+T enc achieved higher estimation accuracy than muscle-length-only estimation, with Holm-corrected p < 0.001.
- Other modality-fusion conditions did not consistently outperform the muscle-length-only condition.
D. Evaluation of Joint-Covering Skin under External Mechanical Disturbances
External disturbances were applied around the elbow to test whether joint-covering skin affects proprioceptive estimation under contact. Errors increased for both modalities, but muscle-based estimation remained relatively stable compared with skin-based estimation.
- External stimulation: External stimulation was applied to the elbow with a stick or by hand during robot motion, producing disturbance data for evaluation.The procedures used approximately 15 minutes of sensorimotor data collection under pressure from various directions.
- Evaluation procedure: Models trained without disturbances were evaluated on sensor data collected under external mechanical stimulation.Single-modality training and estimation procedures matched those used in the preceding experiments.
- Results: Estimation errors increased for both skin and muscle modalities under external disturbances, with particularly large degradation for skin.External contact introduced deformation that differed from the conditions used to train the models.
- Results: The muscle modality retained a mean error of approximately four degrees under disturbance conditions.This indicates that external mechanical stimuli had a relatively limited influence on muscle signals in this experiment.
- Evaluation metric: RMSE values in the external-stimulation comparison were averaged across the three elbow joint angles and paired by random seed over 20 seeds.The comparison included muscle length, Merkel-like skin, and Ruffini-like skin sensing modalities.
V. DISCUSSION
The results indicate that skin deformation provides proprioceptive information complementary to muscle sensing, while disturbance sensitivity and implementation scale constrain the current design. The joint-covering structure may also support disturbance interpretation and mechanical protection.
- Proprioceptive information: Skin-only estimation achieved joint-angle accuracy of approximately three degrees, indicating that deformation around the joint contains proprioceptive information.The result is consistent with biological evidence that skin stretch sensed by cutaneous mechanoreceptors contributes to proprioception.
- Multimodal integration: Encoder-based fusion of muscle and skin modalities outperformed the muscle-length-only condition, suggesting complementary sensory information.The improvement depended on the fusion architecture; other fusion conditions did not consistently outperform muscle-length-only sensing.
- Disturbance interpretation: External contact strongly affected skin estimates because local contact deformation was superimposed on deformation caused by joint motion.This shifted the input distribution relative to models trained under non-stimulated conditions.
- Disturbance interpretation: Comparing skin- and muscle-derived estimates may help indicate external contact or muscle-side abnormalities when the modalities diverge.The proposed interpretation relies on muscle estimates remaining relatively stable in one example and skin estimates remaining stable in the other.
- Broader implications: The joint-covering skin may provide mechanical protection, contact detection, and complementary sensing for robots interacting physically with their environments.The proposed utility extends beyond musculoskeletal humanoids to general robotic systems.
- Limitations: The implementation is a conceptual reconstruction rather than a strict biological reproduction because its sensors are larger and less numerous than biological mechanoreceptors.Future work should improve sensor miniaturization and density.
- Limitations: Long-term stability and high-speed-motion response were not fully evaluated, while hysteresis, fatigue, drift, and quasi-static data constrain current validation.Dynamic-motion evaluation, online calibration, adaptive learning, and time-series models remain future work.
VI. CONCLUSION
The study designed and implemented a tissue-structured biomimetic joint-covering skin for proprioception in a musculoskeletal humanoid. Skin-derived information enabled joint-angle estimation and improved accuracy when integrated with muscle sensing, while the covering may also help interpret and mitigate external stimuli.
- Contribution: The study presented a biomimetic joint-covering skin with a tissue-like structure aimed at enhancing proprioception in a musculoskeletal humanoid.The design integrates numerous receptor-like elements into soft tissue rather than relying on a few high-performance sensors.
- Findings: Skin-derived information improved joint-angle estimation when integrated with muscle sensing.The experiments also demonstrated meaningful estimation using receptor-like elements embedded within the tissue structure alone.
- Implications: The flexible skin covering may mitigate direct transmission of external stimuli to muscles and contribute to interpreting those stimuli.The conclusion frames these functions as possible contributions of the joint-covering structure.