RASSOR: Advanced Surface Operations Robot

• RASSOR is an advanced robotic platform for extraterrestrial excavation that combines dual counter-rotating bucket wheels with adaptive control systems.
• It employs evolved artificial neural network controllers to execute blueprint-guided regolith excavation and enable effective multirobot coordination.
• The system integrates high-fidelity simulation, field testing, and in situ resource extraction methods to support lunar and asteroid surface operations.

The Regolith Advanced Surface Systems Operations Robot (RASSOR) is a specialized robotic platform designed for autonomous excavation and site preparation tasks in extraterrestrial regolith environments, primarily targeting lunar and asteroid surfaces. RASSOR integrates advanced mechanical design—often featuring dual counter-rotating bucket wheel excavators—with autonomous control architectures to address unique challenges posed by the low-gravity, high-uncertainty context of space surface operations. Its development and deployment draw upon a body of research encompassing robotics, evolutionary control, ISRU (in situ resource utilization), terramechanics simulation, and multirobot autonomy.

1. Mechanical Design and Regolith Excavation Principles

RASSOR is characterized by a bucket wheel excavation mechanism. The prevailing design principle, as articulated in the context of asteroid and lunar resource mining, involves a pair of counter-rotating bucket wheels mounted on articulated arms, enabling the robot to maintain traction and stability in environments where gravity provides minimal anchoring force (Nallapu et al., 2017, Nallapu et al., 2017). This design addresses key mechanical challenges:

• Counter-Rotation: By rotating in opposite directions, the wheels cancel horizontal reaction forces, stabilizing the robot and preventing lift-off in microgravity (Nallapu et al., 2017).
• Ballast Mass: Ballast within the wheels increases normal force, enhancing traction on loosely bound regolith (Nallapu et al., 2017).
• Adaptable Geometry: The spacing, sizing, and cutting angle of buckets are optimized for the expected regolith particle size and cohesion, often with geometry parameterized from lunar or asteroid simulant data.

The excavation process models the cutting resistance using soil mechanics tailored for planetary surfaces. For example, the force required to cut through regolith is estimated by superimposing formulations for cohesive and cohesionless soils:

$F_{\text{reg}} = F_{\text{sand}} + F_{\text{clay}}$

where, for example,

$$F_{\text{sand}} = \rho g w l^{1.5} B^{1.73} v d \sin \theta_{\text{eff}}$$

with parameters representing regolith density ($\rho$), gravity ($g$), width ($w$) and length ($l$) of the cutting face, shape parameter ($B$), velocity ($v$), depth ($d$), and effective angle ($\theta_{\text{eff}}$) (Nallapu et al., 2017).

2. Autonomous Control and Evolutionary Behavior

RASSOR's autonomy leverages evolutionary artificial neural network controllers, particularly the Artificial Neural Tissue (ANT) framework (Thangavelautham et al., 2017). In this paradigm, excavation tasks are specified as 3D blueprints, and ANT controllers are evolved to maximize fitness functions that reflect how closely excavation outcomes match the blueprint. The fitness function is quantitatively expressed as:

$$f = \frac{\sum_{(i,j)} \theta_{ij} \cdot \exp(-2|g_{ij} - z_{ij}|)}{\sum_{(i,j)} \theta_{ij}}$$

where $g_{ij}$ is the target depth, $z_{ij}$ is the current depth, and $\theta_{ij}$ is an indicator variable specifying blueprint relevance (Thangavelautham et al., 2017).

Notably, ANT controllers autonomously discover behaviors such as:

• Slot-Dozing: Excavate in controlled passes (incision cuts followed by leveling), often constructing berms for site protection.
• Contingency Handling: Evolved strategies for obstacle avoidance, stuck detection (e.g., by using "rocking" maneuvers), and negotiation behaviors to resolve robot–robot interference.

The decentralized and modular nature of ANT-based control allows each RASSOR unit to adapt to local regolith inhomogeneities and environmental disturbances with minimal human intervention.

3. Multirobot Coordination, Negotiation, and Team Optimization

Multirobot deployments of RASSOR benefit from emergent coordination strategies evolved within the ANT framework. These strategies address challenges such as antagonism (mutual interference when too many robots occupy the same area, risking efficiency loss or gridlock) and dynamic team sizing:

• Sensor Integration: Robots use proximity, force, and depth sensors to detect neighboring agents and modify their trajectories accordingly.
• Negotiation Policies: Evolved behaviors include randomized turning, adaptive back-off, and memory-augmented routines to avoid cyclic conflicts.
• Scalable Optimization: The evolutionary process can concurrently optimize both control parameters and the team size variable to maximize overall excavation throughput (Thangavelautham et al., 2017).

These mechanisms are essential for site preparation tasks requiring rapid progress, such as landing pad construction and the burial of habitat modules for thermal and radiation protection.

4. Resource Extraction, ISRU Integration, and System-Level Operations

RASSOR is frequently conceptualized as the front end of an ISRU chain. Excavated regolith is conveyed for processing, which may entail:

• Thermal Extraction: Heating regolith to $\sim$1000°C liberates trapped volatiles, particularly water, which can be further electrolyzed into hydrogen and oxygen for rocket propellant (Nallapu et al., 2017, Nallapu et al., 2017).
• Energy and Power Modeling: Excavation, heating, and electrolysis subsystems are jointly modeled within the mission's power budget, often constrained to $\sim$10 kW solar input. The dominant consumer is thermal extraction rather than mechanical excavation.
• Fuel Pathways: Both electrolysis (yielding higher exhaust velocities) and direct steam propulsion (trading specific impulse for system simplicity) are analyzed, with RASSOR's continuous regolith collection enabling robust, risk-mitigated ISRU operations (Nallapu et al., 2017).

5. Terramechanics Modeling and High-Fidelity Simulation

Accurate simulation of RASSOR's performance in regolith environments utilizes continuum and particle-based terramechanics models:

• Physics-Based SPH (Smoothed Particle Hydrodynamics): The Chrono::CRM framework employs GPU-accelerated SPH to model interactions between RASSOR's implements and deformable regolith, incorporating plasticity, cohesion, and large-strain behavior. Validation is performed against DEM (discrete element method) simulations and experimental data, showing consistent results for driving torque and soil loading during excavation (Unjhawala et al., 8 Jul 2025).
• Scalable Simulation Domains: CRM's “active domains” technique enables interactive-rate simulation of up to 100 million particles over 10 km terrain (e.g., for mission planning or rover controller testing).
• Accessibility: The CRM simulation toolkit is open source, enabling reproducibility and collaborative research in off-road and extraterrestrial vehicle dynamics (Unjhawala et al., 8 Jul 2025).

6. Field Verification, Deployment Pathways, and Mission Integration

RASSOR development adheres to a structured validation workflow:

• Simulation Phase: Evolved controllers and excavator dynamics are tested in simulated 2D/3D environments using high-fidelity regolith and hardware models (Thangavelautham et al., 2017).
• Controlled Physical Testing: Hardware-in-the-loop experiments with scaled robots in lunar-analog regolith (e.g., domes simulating low light and loose material) are conducted, measuring blueprint conformity and analyzing emergent behaviors such as rocking for unsticking (Thangavelautham et al., 2017).
• Site Preparation and Habitat Construction: RASSOR's capabilities are specifically suited for tasks like landing pad construction (for ejecta protection) and the burial of habitat modules (to maintain subsurface temperatures and shield against radiation). The robots can autonomously interpret excavation blueprints, match surface profiles, and coordinate filling or removal of regolith as needed.

A plausible implication is that RASSOR, equipped with evolved controllers, forms a foundational component in lunar and Martian site preparation workflows, especially in conjunction with swarms of similar robots for enhanced redundancy and throughput (Thangavelautham, 2019, Bier et al., 2021).

7. Comparative Platforms and Broader Context

Relative to other regolith-manipulating systems (e.g., the European Moon Rover System, EMRS), RASSOR is more specialized towards excavation and ISRU-centric roles, while EMRS emphasizes modularity for multipurpose payload accommodation and highly reconfigurable operations (Luna et al., 2023). In large-scale base construction scenarios, autonomous swarms including RASSOR-type vehicles are modeled as achieving substantial improvements in energy efficiency and construction speed compared to human crews, with system-level energy budgets dominated by excavation, transport, and processing tasks (Thangavelautham, 2019).

Summary Table: Principal RASSOR Capabilities and Features

Capability	Design/Control Solution	Research Source
Bucket wheel excavation	Dual counter-rotating with ballast	(Nallapu et al., 2017, Nallapu et al., 2017)
Autonomous, blueprint-driven digging	ANT-evolved neural controllers	(Thangavelautham et al., 2017)
Multirobot coordination	Negotiation, gridlock avoidance, team sizing	(Thangavelautham et al., 2017)
ISRU resource handling	Thermal extraction + electrolysis/steam	(Nallapu et al., 2017, Nallapu et al., 2017)
Physical/procedural simulation	Physics-based SPH, CRM for large terrain	(Unjhawala et al., 8 Jul 2025)
Field deployment pathway	Progressive: simulation → controlled test	(Thangavelautham et al., 2017)

Conclusion

RASSOR exemplifies a convergent approach in lunar and asteroid regolith operations, integrating optimized mechanical design, evolved autonomous control, and scalable multirobot interaction with high-fidelity simulation and a clear validation path toward flight and field deployment. Its foundational role in ISRU, surface preparation, and autonomous construction situates it as a central technology for future off-world infrastructure projects.