On-Body Mechanical Actuators

• On-body mechanical actuators are engineered systems designed to deliver controlled forces to the human body using rigid, hybrid, and soft architectures.
• They integrate compliant materials and programmable kinematics to enable safe and adaptable physical human-robot interaction for rehabilitation, augmentation, and tactile feedback.
• Key design trade-offs, including force output, bandwidth, backdrivability, and wearability, drive advancements in actuator performance and user safety.

On-body mechanical actuators are engineered systems that impart controlled forces and displacements directly to or through the human body in wearable configurations. The paradigm encompasses rigid, hybrid, and entirely soft actuators for physical human-robot interaction (pHRI), rehabilitation, augmentation, haptic feedback, and medical assistance. Modern developments integrate compliant materials, programmable kinematics, and highly customized form factors to address the interplay of safety, performance, adaptability, and interfacing requirements unique to on-body mechanotransduction.

1. Fundamental Architectures and Material Systems

On-body mechanical actuators span a spectrum from rigid-body mechanisms to composite hybrid and fully soft architectures. Hybrid actuators, such as those described by Tang et al. combine 3D-printed rigid shells (exoskeleton, defining revolute joint geometry: $l_r=14$ mm, $\theta=75^\circ$, $r_r=10$ mm, $R_r=12$ mm, $t_r=1$ mm) with serially arranged soft elastomeric bladders (sphere radius $r_s=9$ mm, wall thickness $t_s=1.5$ mm, Shore 28A) to enable mechanically programmable degrees of freedom (DoF) and joint locking by selective pinning (Chen et al., 2020). Purely soft actuators utilize compliant materials (e.g., silicone, TPU) formed into spatially programmable structures such as origami bellows, fluidic fabrics, or multilayered sleeve geometries (Abboodi, 8 Nov 2025, Zhu et al., 2019, Schäffer et al., 2023, Schaffer et al., 2024). Smart materials, including piezoelectric ceramics, carbon nanotube (CNT) yarns/composites, and electroactive polymers (EAPs), are central to tactile and haptic on-body actuators (Xie et al., 2017, Linnander et al., 16 Jan 2026). Rigid or quasi-direct drive (QDD) actuation remains prevalent for high-precision, high-torque exoskeletons, using high torque-density BLDC motors with minimal reduction (Yu et al., 2020, Nesler et al., 2021).

2. Operating Principles and Control Approaches

Rigid and hybrid on-body actuators typically employ revolute or prismatic joints with direct or indirect actuation. Hybrid actuators generate bending by pressurizing internal soft bladders; overall curvature $\kappa$ and bending moment $M$ relate to actuation patterns via

$$\kappa = \frac{\sum_{i=1}^N \omega_i}{N d} \,, \quad M = \alpha\,P$$

where $N$ is the number of active joints, $d$ joint spacing, $\omega_i$ the per-joint deflection, and $P$ the chamber pressure (Chen et al., 2020). Linear hybrid actuators for high-force applications utilize rigidity-constrained soft bladders pressurized within shells, with output force described by:

$F_{\text{out}} = \eta\,P_{\text{in}}\,A$

introducing the efficiency factor $\eta$ ($0<\eta\leq1$) that absorbs compliance and frictional losses (Wan et al., 2020). Soft actuators realize programmable linear, bending, and twisting kinematics by architected pressurization and compartmentalization, as extensively detailed for soft sleeve actuators and fluidic muscle sheets (Abboodi, 8 Nov 2025, Zhu et al., 2019).

Control strategies range from open-loop pressure modulation for monotonic expansion (suitable for timed flexion/extension routines) (Chen et al., 2020) to advanced collocated pressure or endpoint force feedback, as in magnetorheological-hydrostatic actuators, where collocated pressure feedback enables stable, high-bandwidth force control (Véronneau et al., 2022). Backdrivability and impedance are critical: MR-hydrostatic and QDD approaches achieve low output impedance (<2–11% resistive force for MR-hydrostatics, 0.4 Nm backdrive torque for QDD) and high closed-loop bandwidth (e.g., 62.4 Hz for QDD hip exoskeletons) (Yu et al., 2020).

3. Performance Metrics and Comparative Analysis

Bending and Blocked Force:

• Hybrid bending actuators: tip force $\approx4$ N at $P=165$ kPa; bending angles up to $230^\circ$ at $P=120$ kPa; repeatability $\pm5\%$ (Chen et al., 2020).
• Engineered hybrid linear actuators: >100 N at 50 kPa, $\eta$ reaching 97% (Wan et al., 2020).
• Soft sleeve actuators: linear extension force up to 214 N at 200 kPa; bending force 38 N, max bend angle 140° at 200 kPa (Abboodi, 8 Nov 2025).

Bandwidth and Dynamic Response:

• Hybrid actuators (pneumatic): inflation to operating pressure in 0.5–1 s (Chen et al., 2020); soft sleeves, 1–2 Hz for gait cycles (Abboodi, 8 Nov 2025).
• MR-hydrostatic: force bandwidth >25 Hz blocked, 6.5 Hz in compliant loads; rise time down to 56 ms (Véronneau et al., 2022).
• QDD and modular rigid actuation: bandwidths 62.4 Hz, tracking error <5.4% (Yu et al., 2020).

Wearability and Safety:

• Soft actuators: <300 g, sleeve thickness <25 mm, compliant materials (TPU, silicone) with biocompatibility (Abboodi, 8 Nov 2025, Schäffer et al., 2023).
• Rigid/hybrid: edge rounding, low operating pressure, and surface compliance features for skin safety (Chen et al., 2020).
• Passive anchoring strategies (inflated fPAMs): <5 mm displacement at 28 kPa, passive compression $\sim$5 N for limp support (Schaffer et al., 2024).

Architecture	Max Force (N)	Bandwidth (Hz)	Stiffness / Safety
Hybrid (bend)	4 (tip)	~1 (pneumatic)	Programmable via $P$
Hybrid (linear)	100+	20–80 ms rise	$\eta>0.9$ at 50 kPa
Soft sleeve	38 (bend)	1–2	Full-contact, soft
QDD rigid	17.5–30 (Nm)	62.4	0.4 Nm backdrivable
MR-Hydrostatic	25–39 (Nm)	>25	<11% resistive force

4. Design Principles and Mathematical Modeling

Accurate performance prediction for on-body actuators necessitates joint kinematic and mechanical modeling. Hybrid actuators rely on joint-space curvature/pressure models ($\kappa$, $M$); fluidic actuators derive from force-elongation tube models, e.g.,

$$F_{\text{ext}} = N [E\epsilon\pi(r_o^2 - r_i^2) - p\pi r_i^2]$$

for $N$ channels, $E$ modulus, $\epsilon$ strain, $r_o$, $r_i$ tube radii (Zhu et al., 2019). Soft sleeves' orthogonal kinematics (linear/bend/twist) are parameterized via origami fold geometry (e.g., fold angle $\beta$, width $f_w$, wall thickness $w_t$), helical arrangement (twist), and chamber layout (omnidirectional actuation) (Abboodi, 8 Nov 2025).

Efficiency and mechanical loss propagate non-trivially into system dynamics. The apparent inertia with per-joint efficiency $\eta$ is

$$M_{\text{app}}(\eta) = M_{qq} + G^{-T}\, \mathrm{diag}(\eta_i)\,M_{\phi\phi}\,G^{-1}$$

where $M_{qq}$ is the generalized coordinate inertia, $G$ the speed reduction matrix (Sim et al., 2020). Trade-offs between high gearing (large $k_g$), efficiency (low friction), and passive load capacity are governed by closed-form inequalities for static load-sustaining versus dynamic compliance.

5. Application Domains and Integration Strategies

Augmentative and Assistive Exosuits/Orthoses:

• Hybrid and linear compression actuators are integrated into bracelets and straps to correct joint alignment (knee gait assist), via distributed mounting and direct axial loading (Wan et al., 2020).
• Modular soft actuators enable high-torque, multi-DoF exosuits for upper/lower limb rehabilitation and assistance (shoulder exoskeletons with 2-DOF control, wrist fPAMs producing 3.3 Nm at 137 kPa, and lower-limb partial-unloading orthoses) (Natividad et al., 2019, Schäffer et al., 2023, Nesler et al., 2021).
• Fluidic muscle sheets provide area-distributed actuation for compression garments, assistive gloves, and skin-stretch haptic feedback (Zhu et al., 2019).

Tactile and Haptic Interfaces:

• Smart material-based actuators (piezoelectric, CNT, SMA, EAP) enable high-bandwidth, miniature tactile displays, targeting specific mechanoreceptor frequencies; emerging Haptic Light-Emitting Diodes (HLEDs) achieve sub-N force and 1 mm displacement at sub-100 ms latency with direct opto-mechanical drive (Xie et al., 2017, Linnander et al., 16 Jan 2026).

Attachment and Interfacing Components:

• Stretchable pneumatic sleeves (fPAM bands) act as adaptive attachment interfaces, providing passive and active anchoring with low-mount displacement, burst resistance, and user-adaptive comfort (Schaffer et al., 2024).

6. Comparative Insights and Design Recommendations

A central theme is the trade-off between compliance, force capacity, bandwidth, backdrivability, and anatomical conformity. Soft sleeve actuators (force density 2–6 N/cm³, bidirectional/omni-DoF capabilities) excel in adaptability and form-fitting, whereas hybrid and QDD actuators are crucial for precision, large torques, and high control bandwidth. Modularization―mechanical (lockable DoFs, channel patterning) and control (independent pressure regulation, task-invariant assistance)―enables tailored functionality to user and task complexity.

Key guidelines for design include:

• For hybrid-DoF actuators, $P\leq120$ kPa (contact force ≤ 4 N), $d\approx14$ mm for curvature smoothness, and Shore hardness 10–30A for balance between sensitivity and safety (Chen et al., 2020).
• For soft sleeves: operating pressures 100–300 kPa, sleeve thickness <25 mm, geometric tailoring according to joint actuation/reception demands (Abboodi, 8 Nov 2025).
• For tactile actuators: keep skin contact pressure <50 kPa; frequency and amplitude targeting mechanoreceptor bandwidth for perceptual effectiveness (Xie et al., 2017).

7. Emerging Directions and Engineering Challenges

Scalability toward finer spatial resolution (for haptic feedback), lower-voltage/pressure operation (for batteryless/wireless systems), and increased integration of self-sensing (embedded strain/pressure sensing, closed-loop control) drive current research. Fabrication advances (digital sewing, self-healing elastomers, composite materials) are shifting capabilities toward more seamless, robust, and scalable manufacturing, with ongoing studies into bio-compatibility, prolonged wear, and adaptability to user variability.

Rich mathematical modeling of actuator-environment interaction, transmission losses, and human biomechanics continues to inform the balance between mechanical robustness, control performance, and user safety. Modular, reconfigurable architecture and system-level efficiency considerations remain key in next-generation on-body mechanical actuator research and translational deployment (Wan et al., 2020, Sim et al., 2020, Nesler et al., 2021).