SSailOR: Spherical Sailing Rover

• SSailOR is a wind-powered spherical rover that achieves omnidirectional mobility using onboard sails for aerodynamic propulsion.
• It integrates a modular, 3D-printed structure with stepper and servo motor actuation to control yaw and sail trim for efficient wind-driven motion.
• Wind-tunnel experiments validate its co-design approach, identifying optimal yaw and angle-of-attack settings to balance drag and lift performance.

Searching arXiv for SSailOR and closely related spherical rover papers. I’m going to look up the cited SSailOR and related spherical rover papers on arXiv to ground the article in current literature. SSailOR, the Spherical Sailing Omnidirectional Rover, is a wind-powered autonomous rover that uses onboard sails to obtain long-duration mobility with minimal energy consumption. In the formulation reported for wind-tunnel testing, the platform is motivated by applications in planetary exploration, Arctic observation, and military surveillance / reconnaissance, and it combines a spherical design, a main sail, a jib sail, and internal actuation for controlled wind-driven motion (Varanwal et al., 17 Aug 2025). The concept occupies a distinct position within spherical robotics: unlike pendulum-driven or wheel-driven spherical robots, SSailOR is organized around aerodynamic propulsion, yet it retains the spherical rover emphasis on enclosed structure, omnidirectional movement capability, and reduced mechanical complexity.

1. Conceptual definition and operating rationale

SSailOR is presented as a spherical rover that moves using wind propulsion via onboard sails rather than conventional continuous motor-driven traction (Varanwal et al., 17 Aug 2025). The operating principle is based on the interaction of wind flow, the main sail, the jib sail, the yaw orientation of the body, and internal actuation. The reported implementation uses a main sail, a jib sail, stepper motor-driven rotation, a servo-actuated jib sail, IMU feedback, and load-cell-based force sensing. The wind generates aerodynamic forces on the sails, and by adjusting sail angles and rover yaw, the system produces useful drag and lift components that can be exploited for motion and steering.

The spherical geometry is described as enabling omnidirectional movement capability, while also simplifying the mobility architecture relative to multi-wheel or articulated systems (Varanwal et al., 17 Aug 2025). In the paper’s comparison table, SSailOR is identified as both self-righting and upwind-capable. This is a significant distinction within wind-driven robotic mobility, because the platform is not purely passive. A common misunderstanding would be to treat SSailOR as a freely rolling sphere driven only by ambient wind; the reported system instead combines aerodynamic propulsion with active orientation and trim control.

The paper explicitly states that the design requires a co-design approach (Varanwal et al., 17 Aug 2025). In context, this means that mechanical design, sail geometry, actuation, control, and model validation cannot be treated as separable subsystems. A plausible implication is that the rover’s performance envelope is determined less by any single component than by the coupling among aerodynamics, attitude regulation, and spherical rolling mechanics.

2. Mechanical architecture and structural layout

The wind-tunnel article describes SSailOR as a modular experimental platform fabricated entirely by FDM 3D printing and assembled with T-slotted rails, nuts, bolts, and M2 screws (Varanwal et al., 17 Aug 2025). The main structural elements are the traction hoop, curved ribs, side hubs, inner hull, main sail, and jib sail. The ribs connect the hoop to the side hubs using a curved T-frame structure intended to improve stiffness while keeping the body light.

The principal geometric parameters reported for the tested rover are as follows.

Parameter	Value
Rover Diameter	400 mm
Side Plate	19.3 mm height and 174.4 mm diameter
Main Sail Chord Length	100 mm
Main Sail Span	215 mm
Jib Sail Chord Length	70 mm
Jib Sail Span	185 mm
Hull Diameter	215 mm
Ribs dimension	T cross section of 8 mm width
Number of Ribs	8

The hoop was assembled from multiple arc segments joined via interlocking 3D-printed sleeves (Varanwal et al., 17 Aug 2025). The spherical body is supported by a central aluminum rod and uses four bearings to enable low-friction rotation. Rotation is actuated by a stepper motor transmitting torque through a timing belt connected to a driven gear, with a selected gear ratio of 2:1.

The sail geometry is equally specific. The main sail uses NACA 0020, while the jib sail uses NACA 0025 (Varanwal et al., 17 Aug 2025). These are described as symmetric airfoil-like sections chosen to support aerodynamic performance and controlled force generation. The reported architecture therefore integrates spherical rolling structure and sail aerodynamics into a single mechanically coupled body rather than attaching sails to an otherwise conventional rover chassis.

3. Propulsion, control channels, and instrumentation

The propulsion mechanism is organized around force generation on the two sails. The paper interprets longitudinal force $F_x$ as aligned with wind direction and mainly as drag in the test frame, while lateral force $F_y$ is interpreted as lift or side-force contribution (Varanwal et al., 17 Aug 2025). The rover’s motion depends on the balance of these components and on how they vary with yaw angle, sail angle of attack, rotational state (RPM), and wind pressure.

The jib sail is actuated by a servo motor and adjusted using real-time torque readings from the load cell, with the stated aim of maintaining near-zero net torque about the vertical axis (Varanwal et al., 17 Aug 2025). This point is central to the control philosophy. It indicates that the jib is not merely an auxiliary sail for added area; it functions as an actively trimmed element for torque balancing and directional stabilization.

The instrumentation stack was designed for synchronized sensing and actuation. The principal sensors are:

• Six-axis load cell: ATI Gamma FT16284, measuring $F_x, F_y, F_z$ and $M_x, M_y, M_z$
• Hall-effect magnetic sensor: used to verify rotation rate
• IMU: used to track yaw orientation
• Wind sensor: measuring upstream free-stream velocity
• Barometric sensor: recording ambient pressure
• Temperature sensor: included in wind-tunnel flow characterization

The actuation and data-acquisition system combines a micro-stepping driver for stepper motor speed/position control, a stepper motor for rover rotation, a servo motor for jib sail actuation, and NI DAQ modules for analog acquisition and PWM generation (Varanwal et al., 17 Aug 2025). The data architecture has two paths: a microcontroller-based digital acquisition path for IMU, wind sensor, and Hall-effect sensor data, and an NI DAQ-based analog acquisition/control path for six-axis load-cell measurement and PWM generation. LabVIEW served as the operator interface for real-time visualization, sensor monitoring, health display, and actuator-state tracking. All data were logged at 100 Hz.

This instrumentation profile shows that SSailOR, at least in the reported experimental phase, is best understood as a tightly instrumented mechatronic test article rather than a purely conceptual wind rover. The platform was explicitly configured to support aerodynamic system identification and model validation.

4. Wind-tunnel methodology and tested parameter space

Experiments were performed in the subsonic wind tunnel at North Carolina State University (Varanwal et al., 17 Aug 2025). The wind tunnel was used to obtain controlled freestream conditions, repeatable settings, and reduced ambient variability, and to validate aerodynamic and dynamic modeling assumptions. The test section measured

$32 \times 45 \times 46\ \text{inches}$

and, to fit within the tunnel and limit blockage, the researchers used a half-scaled SSailOR model with 400 mm diameter.

The rig consisted of the rover, an upper test mount, and a lower mount. To reduce flow interference from the support hardware, the setup maintained an 80 mm clearance between rover and support structure and incorporated vertical fairings and horizontal fairings. The fairing shapes included NACA 0030 and elliptical fairings, and symmetry was preserved using a dummy motor on the opposite side together with symmetric support structures (Varanwal et al., 17 Aug 2025). These details matter because support-structure wake and rig asymmetry would otherwise contaminate force and moment measurements.

The primary test matrix varied the following parameters:

Parameter	Values
RPM	0, 60, 80
PSF	1.0, 1.5
Yaw	$10^\circ, 20^\circ, 30^\circ, 40^\circ$
Setup Type	Constant RPM, Variable RPM
Main Sail Angle of Attack	$0^\circ, 4^\circ, 8^\circ, 12^\circ, 16^\circ$

In addition to full-system tests, the study also examined ground effect and subsystem configurations (Varanwal et al., 17 Aug 2025). Ground effect was investigated by placing a flat steel plate beneath the rover, supported by four NACA 0025 airfoils. Component-level decomposition was performed using three configurations: structure only, hoop + ribs, and sails + inner hub. These subsystem tests were intended to isolate the aerodynamic contribution of the frame, the hoop/ribs assembly, and the sails plus inner hub.

The paper notes that it references a simplified dynamic model developed previously, but does not provide explicit governing equations in the text shown (Varanwal et al., 17 Aug 2025). The directly reported measured quantities are the six load-cell outputs $F_x, F_y, F_z, M_x, M_y, M_z$, interpreted in a wind-aligned frame with $X_0$ as drag direction and $Y_0$ as lift direction.

5. Experimental observations and aerodynamic behavior

The wind-tunnel results identify several stable qualitative trends. For longitudinal force / drag $F_y$0, the paper reports that $F_y$1 was generally negative, consistent with the chosen sign convention for drag (Varanwal et al., 17 Aug 2025). The magnitude of drag increased with wind speed and yaw angle, with the largest drag values occurring at Yaw = 30° and 40°, especially at 1.5 PSF. Across sail angles of attack, drag was described as relatively flat at lower yaw angles and slightly reduced near 0° to +8° AoA in some cases.

For lateral force / lift $F_y$2, the paper reports stronger dependence on sail angle of attack (Varanwal et al., 17 Aug 2025). Lift was maximized around Yaw = 20°, AoA near +8°, and higher wind pressure and RPM. The reported peak lift reached about 2 N. At Yaw = 40°, lift was reduced and sometimes negative.

These observations support a specific interpretation of SSailOR’s aerodynamic coupling. The rover does not respond to sail angle alone; effective propulsion requires joint optimization of yaw, main sail angle of attack, wind condition, and rotational state (Varanwal et al., 17 Aug 2025). This directly undercuts another likely misconception, namely that the vehicle can be steered or optimized by sail trim independently of body orientation. The reported data instead indicate a multi-parameter operating envelope in which yaw-sail coordination is decisive.

Ground-effect testing further showed that proximity to the ground can alter force and torque measurements, although the summary does not report exact numerical conclusions (Varanwal et al., 17 Aug 2025). The subsystem tests likewise indicate that structural drag from the frame, aerodynamic contribution of the hoop/ribs, and the contribution of the sails and inner hub must be separated analytically if the design is to be optimized in a principled way.

The paper treats these results primarily as validation data for dynamic-model refinement. The measurements support model assumptions that aerodynamic loads depend on sail orientation and vehicle yaw, that wind speed strongly affects force magnitude, and that there is coupling between internal actuation and aerodynamic response (Varanwal et al., 17 Aug 2025). At the same time, the results show that any predictive model must account for nonlinear yaw effects, sail-angle sensitivity, configuration-specific aerodynamic interactions, and ground effects.

6. Position within spherical rover research

Within the broader arXiv literature on spherical robots, SSailOR belongs to a family of platforms that use spherical geometry to address mobility, protection, and terrain adaptability, but it differs sharply in propulsion modality. SphereX, for example, is a 2 kg, spherical, holonomic planetary robot with a hopping mechanism for rugged terrain traversal and deployment from a larger rover, targeting locations such as slopes, canyons, cliffs, crater rims, and pits (Raura et al., 2017). In that lineage, the spherical body supports tactical access to hazardous terrain through rolling and hopping rather than wind propulsion.

A separate line of work examines spherical robots as mobile mapping platforms. The paper on two spherical systems for mobile 3D mapping reports both a lightweight non-actuated sphere and an actuated pendulum-driven sphere, each carrying a Livox Mid-360 LiDAR and running FAST-LIO2, DLIO, and FAST-LIVO2 in LIO-only mode on a Raspberry Pi 5 (Khalil et al., 12 Sep 2025). The main result is that the performance of state-of-the-art LIO algorithms deteriorates under spherical rolling motion, producing globally inconsistent maps and sometimes unrecoverable drift. This is relevant to SSailOR because it suggests that spherical mobility creates perception and state-estimation challenges even when the propulsion mechanism is not aerodynamic.

Another comparison point is RotunBot, a spherical mobile robot derived from a single-pendulum-driven spherical robot but extended with a momentum wheel aligned with the pendulum’s lateral-swing axis (Zhang et al., 3 Nov 2025). The paper reports stable high-speed motion at up to 10 m/s through simple decoupled control, together with enhanced obstacle-crossing performance and terrain robustness. RotunBot therefore addresses a different systems problem: stabilizing a mass-shifting spherical robot at high speed by decoupling attitude setting from roll-rate stabilization.

Taken together, these studies clarify what is specific about SSailOR. Unlike SphereX, it is not centered on hopping or deployable micro-rover tactics. Unlike pendulum-driven mapping spheres, it is not primarily a LiDAR carrier with internal mass-shifting locomotion. Unlike RotunBot, it does not derive its mobility from pendular drive plus momentum-wheel stabilization. SSailOR instead defines a wind-propelled branch of spherical rover research in which aerodynamic force production, spherical rolling, and active trim/orientation control are inseparable design variables (Varanwal et al., 17 Aug 2025). This suggests that its most consequential research challenges lie at the intersection of unsteady aerodynamics, attitude-control allocation, and terrain-proximate rolling dynamics.

7. Design implications, caveats, and research outlook

The reported SSailOR study is explicitly an experimental setup and validation study rather than a field-deployment report (Varanwal et al., 17 Aug 2025). Its contribution is the design, instrumentation, and test procedure for evaluating the rover in a controlled wind-tunnel environment, together with observations relevant to further optimization. The main design implication stated in the paper is that the rover requires a co-design approach integrating sail geometry, yaw control, support structure and fairings, ground interaction, and control strategies based on real-time torque and orientation feedback.

The paper identifies the best observed operational regions as generally around Yaw 10°–20°, AoA around +8°, higher wind availability, and careful tuning of RPM and sail trim (Varanwal et al., 17 Aug 2025). This should not be read as a universal operating law; it is an experimental observation specific to the reported setup, wind-tunnel conditions, and half-scale rover. A plausible implication is that SSailOR’s future development will depend on whether these operating regions remain favorable under off-rig conditions, terrain interactions, and outdoor turbulence.

Several caveats follow directly from the reported methodology. First, the tests were conducted on a half-scaled model in a wind tunnel, with deliberate efforts to mitigate support-structure interference (Varanwal et al., 17 Aug 2025). Second, ground effect was treated as a variable of interest, indicating that near-surface flow modification is not secondary but structurally relevant to rover behavior. Third, the paper does not provide, in the summary shown, explicit governing equations for the full dynamic model, which means the experimental data currently serve as a primary basis for model refinement rather than as a final validated performance law.

In the context of spherical rover research, SSailOR therefore represents a specialized but coherent architecture: a spherical, sail-driven, omnidirectional rover whose viability depends on experimentally grounded integration of aerodynamic design, actuation, sensing, and control (Varanwal et al., 17 Aug 2025). Its significance lies less in a claim of mature autonomous deployment than in establishing a testable framework for wind-powered spherical mobility and in showing that aerodynamic performance, unlike in many conventional rover designs, cannot be decoupled from the platform’s global mechanical and control architecture.