ABSTRACT. In this paper, we propose a new initiative for the ISTVS Wheel/Track Design Student Competition, aimed at engaging students in terramechanics and providing hands-on experience in off-road wheel/track design and testing. The competition will challenge student teams to design and fabricate a wheel or track system for a small vehicle, which will then undergo a series of tractive performance and mobility tests in selected soil types. By minimizing costs and focusing on practical, accessible design and testing methods, the competition ensures that students from various backgrounds and disciplines at any interested university can participate.

The competition will include design presentations and performance tests, evaluated using structured scoring criteria. To support accessibility, teams can use freely available CAD software, 3D printing, and other rapid prototyping techniques.

The performance evaluation will be conducted using a test matrix that examines both structural integrity and performance metrics, which may include slope tests, measurement of drawbar pull, sinkage versus slip, lateral forces in turn, and slope climbing characterization in alignment with ISTVS standards. A single-wheeled terramechanics test rig will assess the wheel/track performance in controlled environments, optionally supplemented by the small-vehicle based mobility tests under various soil conditions in the field. Through this competition, students will gain valuable experience, develop innovative solutions, and develop connections to ISTVS, fostering the next generation of terramechanics engineers.

ABSTRACT. The development and operation of exploration rovers is essential for human lunar exploration, which is considered an international space exploration goal. However, the lunar surface is a completely different environment from that on the Earth because it is covered with regolith. Hence, it is difficult to evaluate rover’s traveling performance in advance. Under these circumstances, the extended terramechanics (xTerramechanics) model has been proposed that takes terrain surface deformation into account. Meanwhile, in a wheel traveling analysis, stress evaluation models are important in the evaluation of wheel-ground interactions such as forces, torques, and soil movements. In addition to the well-known parabolic stress evaluation model proposed by Wong and Reece, Tsubaki and Ishigami have proposed a Gaussian stress evaluation model based on direct measurement of stress distributions. This model may be able to suppress excessive shear stress behind the wheel and predict more realistic traveling performance even in assuming regolith surface. Therefore, this study systematically examined the effects of these stress distribution models by conducting a traveling analysis using xTerramechanics model. The results show that for small-diameter wheels, the Gaussian distribution model can predict practical traction performance with fewer parameters.

ABSTRACT. Recently, moving underground has been focused on as a method of planetary exploring robots' movement.　Exploration rovers have difficulty moving underground because the drag force from the ground restricts their movement, hindering underground exploration. A previous study developed a mechanism that moves underground by imparting vibration to the ground and decreasing the drag force. The experimental results demonstrate the superiority of the proposed method, as the movement distance achieved with vibration is considerably greater than without vibration. The optimal vibration should be selected to improve the proposed mechanism's underground mobility. When imparting vibration, the change of the bearing capacity, which is the resistance from the ground, depends on lift-off acceleration. Lift-off acceleration is the minimum vibration acceleration that causes the flow of the ground particles. The bearing capacity increases when the rod's movement is under lift-off acceleration. Conversely, it decreases when the rod's movement is over lift-off acceleration. If the simulation reproduces the change of the bearing capacity when imparting vibration and lift-off acceleration, the optimal vibration can be selected. This study conducted the experiment and the DEM simulation in which the rod is dragged while vibrating on the ground. These methods measured the bearing capacity that the rod received from the ground. In the experiment and simulation, the vibration frequency was changed. As the vibration frequency increases, the acceleration of the rod's movement caused by the vibration eventually exceeds the lift-off acceleration. The reproducibility of DEM simulations was evaluated by comparing the relationship between the bearing capacity and vibration frequency in each method.

ABSTRACT. Planning of lunar construction projects, such as the Artemis Base Camp, is challenging. It includes deciding the design and numbers of different construction machines and how to coordinate them to reach certain time and energy budgets. The machines need to be unmanned to avoid human exposure to the harsh lunar environment. They also need to be highly autonomous due to the poor and delayed communication. Simulation has a central role for finding the necessary solutions for all this.

We present a simulation framework for lunar construction work involving multiple autonomous machines. The framework supports modelling of construction scenarios and autonomy solutions, execution of the scenarios in simulation, and analysis of work time and energy consumption throughout the construction project. A behavior tree manages the operational logic and error handling. High-level decision-making is separated from lower-level control algorithms, with the two connected via ROS2. Machine movements are planned using inverse kinematics. The simulations are based on the physics engine AGX Dynamics that integrates contacting multibody dynamics with deformable terrain.

We present results from a test case that involves an excavator and a dump truck that jointly remove lunar soil to construct a ground shape with a specified 3D profile. The excavator switches between the tasks of digging material from the target area and dumping it on the dumptruck, which has the tasks of receiving material, drive back and forth between the excavation area and dump area, dumping material, and doing levelling at the excavation area usings a blade. All these tasks are in turn be broken down into subtasks, that rely on sensors and controllers. The whole system is represented and managed by a behaviour tree.

ABSTRACT. The ability of compliant tires to deform plays a crucial role in enhancing their obstacle negotiation capability compared to rigid wheels. This work presents an efficient method for evaluating the effect of tire deformation by adjusting the contact parameters of a rigid wheel model. Instead of explicitly solving for structural deformations, the approach incorporates constraint kinetic energy at the moment of contact to account for deformation effects, ensuring computational efficiency while maintaining accuracy.

The method dynamically modifies contact parameters, such as normal vectors and contact points, to reflect the altered interaction between the tire and the obstacle. This adjustment enables the introduction of a reduced model that efficiently incorporates the effects of tire deformation without explicitly modelling structural flexibility, providing a computationally efficient yet accurate approach to analyzing obstacle negotiation.

The simulation results demonstrate that the proposed approach effectively replicates key deformation-induced behaviours observed in high-fidelity models while achieving real-time performance. This makes it a practical and efficient solution for vehicle dynamics studies, robotics, and autonomous vehicle applications where both computational efficiency and physical accuracy are essential.

ABSTRACT. Accurate modeling of tire–terrain interaction is critical for effective vehicle dynamics simulations, particularly in off-road environments. A previous study demonstrated the feasibility of generating three-dimensional (3D) terrain models from images captured by an unmanned aerial vehicle (UAV) and identified several areas for further investigation—chief among them the influence of UAV flight altitude and image overlap ratio. This study investigates these parameters, along with the impact of the interpolation technique used to regularize point cloud data.

New UAV imagery of a Belgian paving test track was collected, and baseline measurements from a traditional mechanical profilometer were used for direct comparison. While constructing 3D terrain models from UAV data is relatively straightforward, our findings highlight several critical factors – such as flight altitude, image overlap, and interpolation method – that can significantly affect model accuracy. Based on these results, we propose a practical, step-by-step approach for developing accurate terrain models suitable for use in vehicle dynamics simulations.

The UAV-based technique is broadly applicable, from smooth man-made surfaces (e.g., ISO 8608:2016 Class C roads) to rough, natural off-road terrains, enabling new opportunities for mobility research across a wide range of challenging environments.

ABSTRACT. An important factor in route planning for autonomous off-road vehicles is determining soil condition, particularly load-bearing capacity. Direct (in situ) measurement of this soil property is challenging, and its determination through remote sensing technologies is an emerging area of research. In laboratory tests using prepared soil samples, a strong correlation was observed between the soil surface reflectance in the visible range (400-700 nm) and the load-bearing capacity. A bevameter device was used to determine the load-bearing capacity of each sample at various moisture contents. Measurements were taken for eight different soil textures at different moisture levels, starting from the dry state and gradually increasing to full saturation. This approach allowed us to observe how changes in moisture content affect the load-bearing capacity of each soil texture. Reflectance measurements were then performed on these samples using a portable spectrophotometer. The optical reflectance data were compared with the load capacity measurements obtained from the bevameter. This process enabled the development of a predictive equation that incorporates constants representing the soil's reflectance in its dry state. The resulting equation accurately captures the relationship between reflectance and load-bearing capacity under laboratory conditions. While these results are based on a controlled environment, they offer a valuable foundation for estimating load-bearing capacity in real-world scenarios. Further research, incorporating field data and accounting for environmental variables, will be essential for refining this approach and developing more accurate predictive models for autonomous off-road vehicle navigation.

ABSTRACT. Cold regions represent an especially challenging environment for maneuver. In addition to standard obstacles which occur in less extreme environments, the surface itself can become an obstacle. Winter surfaces are characterized by low strength, resulting in deep vehicle sinkage with high motion resistances. In this work, we developed a cold regions route planning (WRP) algorithm that identifies low risk routes that minimizes the energy required by the vehicle to complete a mission. The algorithm fuses weather, terrain, and vehicle information to generate a unique on and off-road graph of the area of interest. A winter surface algorithm is used to predict the current snowpack’s depth and density using weather and terrain information. This information along with vehicle data, is used to generate the vehicle’s motion resistance and work associated with surface compaction, tree push over force, and gravity forces. Search requests are initiated, and optimal routes are displayed through kml files, enabling the use of existing GIS software to interface with the WRP algorithm. Experiments were conducted at the Camp Ethan Allen Test Site to validate the algorithm. Driven and predicted routes and mission completion times were compared with favorable results. Additional algorithm enhancements are identified to improve route accuracy and in-field operation.

ABSTRACT. The RT3 Simplified Curve Unit (SCU) friction tester, by Halliday Technologies, measures surface friction by operating the test wheel at a small side-slip angle. It can be operated at highway speeds and can be used with either aircraft or truck tires and was specifically designed for robust operations in cold environments in the far north. The US Army Ground Vehicle Systems Center (GVSC), Ground Vehicle Power and Mobility, Tires Track & Suspension Team purchased the Halliday RT3 SCU (SCU) friction tester for use at the Arctic Regions Test Center (ARTC) in Alaska. Before sending it to Alaska, the SCU was evaluated for comparison to standard friction test techniques and cold room performance at the Cold Regions Research and Engineering Laboratory in Hanover, NH. On showing good performance at temperatures down to -45°C (-50°F), and excellent correlation with standard friction test techniques over a wide range of friction surfaces, it was shipped to Alaska and successfully used on snow and ice test surfaces during March and April of 2025.

This paper presents how the SCU tester measures and reports friction values, test results for the SCU unit and other friction measurement techniques on a wide variety of on- and off-road surfaces, the results of cold room testing, and the operational and exploratory measurements on snow and ice fields in Alaska. A discussion of the measurement technique and results is presented along with conclusions for optimizing use on arctic terrain surface characterization and monitoring for vehicle and tire performance testing.

ABSTRACT. The traction-longitudinal slippage curve characterizes the interaction of a driving wheel and deformable terrain and has important considerations affecting the performance of the wheel and of the vehicle. Knowledge of the slippage is crucial to off-road vehicle mobility and energy efficiency assessment and wheel control applications. To be able to analyze and compare the efficiency of different tires on the same terrain, or a tire on different terrains, the tire slippage should be split into two components to determine contributions of the tire and soil to the tire traction characteristic. A new exponential formulation of the longitudinal slippage was previously derived for this application. Many analytically-inspired-empirical based math models exist to describe slippage curves, which are validated, recommended, and being used in software products to approximate tire-terrain performance. To use these models in research with split tire and soil contributions would require rigorous research effort on re-deriving equations for each model because each is based on different mathematical theory and statistical constraints; such effort may not lead to a positive result. As an alternative approach, the analytical split traction-slippage curve is proposed as a general model to approximate other formulations of the slippage curve which can then be split into separate tire and soil slippage curves. Computational simulations of other slippage models are performed and their resulting curves are fit to the new exponential function to analyze the effectiveness of the split slippage model to approximate other models of traction-slippage curves and soil characterization. Using the split traction model, equations are derived to quantify tire and soil contributions the slip power losses.

ABSTRACT. The article will present a system that allows automation of the process of performing terrain passability analysis. The use of automation allows this analysis to be performed quickly and in any area of interest. As part of the research, methodologies have been developed for determining the index of passability (IOP), the task of which is to calculate measurable conditions of vehicle passability in a specific area, taking into account a variety of factors. The main task of these methodologies is to integrate data from various sources (mainly geospatial data: digital maps, meteorological and soil data) and determine an easy-to-interpret coefficient that will additionally take into account the traction characteristics of military vehicles. The determined coefficients will allow the generation of easy-to-interpret passability maps. Combining such a large amount of data from different sources made the resulting maps change their content over time as the meteorological situation changes. The terramechanics experiments carried out, made it possible to take into account the ability of specific vehicles used by the armed forces to travel on roads and roadless areas under specific meteorological and hydrological conditions. Taking into account the specifics of the weather data, the system generates passability maps both taking into account the current situation and also allows for the prediction of passability conditions. In this way, an “on-call, automatically updating passability map” was created, this allows the system to provide the commander with information on the ability to move troops not only for the present moment, but also, taking into account the forecasts. This gives an entirely new capability for planning the movement of troops over roads and roadless areas.

ABSTRACT. Emergency driving scenarios require combined slip, where tyres generate longitudinal and lateral forces to slow the vehicle while following a path. Stability systems e.g. ABS brakes and Electronic Stability Control assist in regaining control after loss of stability. These systems do not actively prevent instability. ABS and ESC systems do not have direct rollover prevention, which is necessary in vehicles with a high centre of mass. For off-road vehicles, this challenge is exaggerated by uneven terrain, loose soil, and rollover risk. Stability control systems struggle with unpredictable road inputs, often compromising handling. A preventative system is needed that manipulates load transfer to improve tyre contact before traction limits are exceeded, and where the roll angle of the vehicle can be directly controlled. This study proposes using a semi-active suspension system to improve performance during combined slip manoeuvres. By adjusting suspension parameters, the system manipulates load transfer and enhances overall traction limits, reducing instability and rollover risk while maintaining steering and braking. A vehicle, equipped with a controllable semi-active suspension system, is used as a test platform. A high-fidelity model is updated and used for parametric studies. Results are validated using experimental tests, with open-loop manoeuvres using brake and steering robots for repeatability and controlled assessment. Preliminary findings indicate that a semi-active suspension system can manipulate combined slip performance, with potential to reduce instability and rollover. While this research is performed on smooth road surfaces, it provides a critical foundation for off-road applications, where terrain variability further challenges the vehicle’s stability.

ABSTRACT. This research considers simulation-based optimization of a tracked vehicle suspension system. We focus on maximizing ride quality across a large spectrum of operational conditions rather than specific terrain/speed combinations, using surface roughness to guide our exploration of diverse scenarios.

The vehicle suspension system features 14 torsion rods and dampers subject to optimization. Using multibody dynamics simulation, we evaluate ride quality over diverse terrains at varying speeds, computing whole-body acceleration and vibrational dose values (VDV).

We analyze and compare alternative ways of posing and solving the problem of finding a suspension configuration that is optimal for a desired spectrum of terrains. We consider derivative-free optimization methods, including Particle Swarm Optimization (PSO) and Bayesian Optimization, and compare these to first learning and pre-building a surrogate model, prior to optimization. This surrogate strategy relates local terrain characteristics to vehicle performance metrics, such as acceleration-constrained velocity limits.

Preliminary work with PSO demonstrated significant ride quality improvements, with VDV reductions of 20-60% compared to baseline configurations, but was limited to optimizing only torsion rods for specific terrain/speed combinations. Our current research extends this by simultaneously optimizing both suspension and damping parameters and implementing an aggregated approach that optimizes performance across multiple terrains and velocities in a single cycle.

This work demonstrates how simulation-based optimization techniques can significantly reduce whole-body vibrations in tracked vehicles, improving both operational capabilities and crew endurance in challenging environments.

ABSTRACT. Understanding the behavior of granular materials under various low-gravity conditions is essential for planetary exploration, as it impacts lander interactions, rover mobility, and in-situ resource utilization processes. However, ground-based experiments using parabolic flights and drop tower facilities are limited by short-duration microgravity conditions, making it challenging to fully capture behavior of granular materials. To address this, numerical simulations provide a powerful alternative. This study employs the discrete element method (DEM) to simulate sand-outflow behavior from hoppers under reduced gravity, based on the Hourglass experiment conducted aboard the International Space Station (ISS). The experiment utilized artificial gravity (0.06G–2G) to observe the deposition and flow characteristics of different granular materials. Using the DEM, we modeled silica sand samples (#5 and #8), incorporating contact forces, rolling resistance, and van der Waals forces/electrostatic forces to evaluate their impact on particle motion. The results show that silica sand #5 exhibited minimal adhesion effects, whereas silica sand #8 demonstrated significant accumulation of sand on walls, particularly under lower gravity conditions. Moreover, neither van der Waals nor electrostatic forces alone could fully replicate experimental results, indicating that both must be considered for accurate modeling. These findings emphasize the importance of interparticle forces in regolith behavior under low gravity and provide insights into granular material interactions in extraterrestrial environments. This study could contribute to the design and optimization of robotic systems for planetary exploration, particularly in excavation, mobility, and construction applications.

ABSTRACT. This paper describes a method to clarify the interaction between wheels and regolith particles when a rover runs on the lunar surface. The regolith dust that lifts during movement can adhere to the equipment mounted on the vehicle, potentially causing significant impacts on its functionality. To mitigate the effects of regolith dust on the onboard equipment, it is crucial to understand the behavior of regolith particles that are lifted by the wheels. Additionally, quantitatively understanding the interaction between wheels and regolith particles is important because it affects the wheels' traction and the vehicle's running performance. Many experimental and analytical studies have been published to address this. In the case of numerical analysis, studies using Discrete Element Methods (DEMs) to clarify microscopic interactions have been conducted. To represent the behavior of regolith particles in numerical analysis, it is necessary to identify many parameters that characterize the physical behavior of particles based on test results, which can be time-consuming due to trial and error in deriving these parameters. In this study, a new data-driven set-based searching method, called Differential Evolution Based on Adaptive Sampling (DEBAS), is adopted for the parameter identification of the particles. The proposed method can efficiently find combinations of particle characteristic parameters that match the desired element tests. The parameters obtained through this method will be verified by comparison with several characteristic tests. Furthermore, obtained parameters are validated by comparison with a scaled wheel-regolith simulant interaction test.

ABSTRACT. Purpose: The impact of gravity and the use of gravity offload on rover tractive performance tests is evaluated and compared between Earth and lunar environments. Single-wheel experiments and DEM simulations of a micro-rover wheel are conducted in Toyoura sand and FJS-1 lunar regolith simulant. Methods: Slip conditions, traction coefficient, and grouser pitch are measured for single-wheel tests across fixed slip values at a wheel load of 24.5 N and a translational speed of 0.02 m/s (low-speed regime). Simulations, previously verified against experiments, are then used to predict lunar gravity performance. Results: Wheel performance in traction coefficient and efficiency under lunar gravity aligns with Earth gravity for the same mass, as does granular soil motion. However, tractive force and resistive torque scale with the lunar-to-Earth gravity ratio (1/6). Earth tests with a 4.1 N wheel load over-predict the traction coefficient relative to lunar conditions if the soil is not gravity-scaled. Applying a 24.5 N load in lunar gravity has a significant impact on soil motion. Conclusion: Gravity offloading alone is unsuitable. Depending on the performance aspect studied, alternatives include Earth tests at full rover mass with no soil scaling, applying granular scaling laws, or reducing the rover scale. This may change with a transition to high-speed wheel motion.

ABSTRACT. The wheeled rover has played an important role in planetary or lunar exploration. The planet and moon surfaces are covered with granular materials, which causes the wheeled rover to get stuck. To suppress wheel slippage, the rover, which utilizes shear force beneath the anchored wheel, can effectively climb a loose surface with steep slopes. Previous studies have developed various types of push-rolling and wheel-walking rovers. Furthermore, our previous study developed a minimal configuration rover and confirmed that it can climb loose surfaces with steep slopes while suppressing slippage. During its movement, the rover changes its wheelbase length and moves its wheels using the resistance forces of the anchored wheels. However, the variable speed of the wheelbase length remains constant at 20 mm/s, and the effect of this speed on the rover’s traveling performance is still unclear. The traveling speed of the rover with small wheels tends to slow down; therefore, understanding the potential to improve its speed is important. The purpose of this paper is to validate the impact of the changing speed of the wheelbase length on the climbing ability. This paper presents a testbed rover that can change the speed of the expanding/contracting wheelbase length and performs the climbing experiment on loose soil with steep slopes. The experimental results suggested that driving performance was maintained even when the expanding/contracting speed increased three times faster than the conventional speed.

ABSTRACT. Tracked vehicles are commonly used on construction sites because of their high vehicle stability and traversability with large ground contact area. Grousers attached to the track shoe enhances vehicle stability and traversability, mitigating the risk of slipping on rough terrain. While many past studies investigated the mechanical effect of a single grouser shoe, experimental test of an entire grousered-track unit remain limited due to non-uniform the stress distribution of the track unit on the ground. In this study, we developed an experimental apparatus with a grousered-track unit and conducted drawbar pull experiment on sandy terrain. We measured the slip ratio and sinkage of the track unit with various traction load corresponding to the drawbar pull of the track unit. The relationship between the drawbar pull coefficient, which is the ratio of the drawbar pull to the vertical load of the track, and the slip ratio was examined. The experiments confirmed that the slip ratio remains constant when the drawbar pull coefficient is approximately 0.4 or lower. Within this range, the track unit maintains small sinkage and a stable driving state. However, when the drawbar pull coefficient exceeds approximately 0.45, the slip ratio sharply increases as larger traction load cases the track unit to an unstable driving state by repeating the move-and-stop motion. This trend was observed regardless of the varied vertical load. These findings suggest that the relationship of the slip-drawbar pull coefficient and the threshold for stable/unstable driving state are useful for evaluating the traversability of the grousered-track unit.

ABSTRACT. The prediction of vehicle performance during off-road missions has historically been a complex task due to the interaction between tyres and soft soil. A comprehensive characterisation of vehicle performance across different terrains is essential for a full understanding of tractive and handling behaviour. Moreover, it is crucial to discuss the tools and methods needed to extract relevant information from off-road testing, as standardisation in this area is lacking compared to on-road testing. This paper focuses on characterising vehicle performance using experimental data collected during an off-road testing campaign conducted across various terrains, including Kalahari Desert sand and gravel pavements. For longitudinal dynamics, coast-down, acceleration, braking, and Drawbar Pull tests are performed to assess rolling resistance, traction, and braking capabilities under off-road conditions. For lateral dynamics, constant-radius tests are conducted on gravel terrain to evaluate cornering behaviour. The synthesis of test analysis results serves as the foundation for developing simplified lumped-parameter handling models and estimating contact parameters for tyre–terrain interaction models.

ABSTRACT. Frost susceptible soils include a wide variety of soil types from highly organic to fine grained silty soils. These soils can pose significant mobility challenges during each of the four seasons especially during the freeze/thaw transition periods between seasons. These mobility challenges have been on display during several conflicts throughout history with the most recent being the war in Ukraine. Maintaining mobility superiority during conflict is critical to avoiding immobilization and becoming a target while maintaining mobility in the Arctic can have similar consequences because immobilization can mean the difference between life or death due to the harsh environmental conditions. This research studies the effects on vehicle mobility across three vehicle classes including both wheeled and tracked platforms operating on peat and silty soils at field capacity. Testing occurred on unfrozen, partially frozen, fully frozen, and partially thawed soil conditions to capture the maximum tractive force during these transitional seasons and better predict cross-country vehicle mobility.

ABSTRACT. The traditional semi-empirical terramechanics models have been heavily relied on over the past decades. Despite the fact that their numerous assumptions limit their applicability in certain scenarios. In particular, the real forces in transient states can results in behaviour that is not captured by the semi-empirical models. Additionally, the models are formulated with the assumption that the wheel or track is travelling over flat, level terrain. We have previously proposed two separate methods for adapting the models to address these two specific shortcomings. For the former, we proposed a hybrid modelling approach which uses trained neural networks to augment the force predicted by the semi-empirical models. Using this approach, we do not discard the valuable insight that the established models provide, but are still able to capture some dynamic and transient behaviour. For the later shortcoming, we proposed a novel variable terrain height adaptation of the semi-empirical model which computes stresses at each point along the wheel-soil interface by projecting that point in space onto the heightfield that defines the terrain. This is in contrast to other established methods for handling uneven terrain by defining least squares planes which are meant to approximate the terrain geometry, but which in practice can have the effect of smoothing out terrain roughness. In order to validate these two proposed models, an experimental test campaign was carried out in which an instrumented off-road vehicle underwent a drawbar-pull test on soft terrain, and a second test in which the same vehicle was driven over a sand course with an uneven surface profile. The results of these two sets of experiments are used to justify the modelling approaches we have proposed.

ABSTRACT. The need to move vehicles off-road arises primarily in military operations and crisis management, where standard road networks may be insufficient or inaccessible. While vehicles typically follow existing roads, critical situations often require off-road movement to execute surprise maneuvers or reach remote locations beyond the infrastructure network. In such scenarios, effective planning is crucial and requires detailed geographical data to assess terrain conditions and vehicle mobility. To support this process, terrain passability maps are essential. These specialized maps synthesize geospatial data, classifying terrain based on how different types of vehicles can navigate it. Despite advancements in vehicle traction technology, such as improved suspension, tires, and tracks, many vehicles still struggle on muddy or waterlogged soils. The war in Ukraine underscored this issue, with reports of modern Russian tanks failing to traverse swampy, clay-rich terrain. This research presents advanced methods for dynamically determining off-road routes with varying difficulty levels while accounting for changing weather conditions. The proposed solutions not only enable real-time route adjustments based on current weather patterns but also incorporate meteorological forecasts, allowing for more strategic long-term route planning. The route determination process relies on the creation of dynamic terrain passability maps, which serve as a foundation for generating a graph that calculates optimal paths. These maps take into account the impact of precipitation on soil properties. By integrating this data, routes can be precisely adapted to both environmental conditions and terrain characteristics, significantly improving efficiency and safety in difficult operational areas.

ABSTRACT. Terrain passability assessment is a crucial aspect of military operations, influencing strategic mobility and tactical decision-making. Traditional methods of evaluating terrain traversability rely on static maps and expert judgment, often lacking the precision needed for dynamic environments. This paper presents an AI-driven approach to terrain passability modeling using artificial neural networks (ANNs). In research two models ann_full and ann_fast— were developed to predict passability coefficients based on geospatial and hydrological data. The ann_full model utilizes a comprehensive dataset of terrain attributes, while the ann_fast model focuses on the most influential parameters to enhance computational efficiency. The models were trained and validated using military geospatial databases, including VMAP Level 2, applied to the Suwałki Gap region. Results indicate that the ANN-based models outperform conventional methods by providing a more granular and accurate assessment of terrain conditions. Comparative analysis of different modeling techniques, including statistical threshold methods and vegetation roughness factor (VRF) models, further underscores the advantages of machine learning approaches in terrain analysis. The proposed AI-driven methodology enhances decision support systems, providing real-time, data-driven insights for military and autonomous vehicle applications.

ABSTRACT. This paper presents the development of a physics-based brush tyre model extended for applications beyond conventional smooth surfaces. The research addresses limitations in existing tyre models when simulating vehicle dynamics on non-smooth, hard terrain. A comprehensive literature survey informed the methodological approach, which combines a physics-based brush model with empirical tread and soil models implemented in MATLAB. Initial validation results demonstrate promising performance, indicating that the combination of empirical soil characterization with the physics-based brush model effectively predicts tyre forces on non-smooth terrain. The model provides a framework for predicting traction limits in off-road conditions, expanding the simulation capabilities for vehicle dynamics on complex surfaces.

ABSTRACT. Modelling flexible tracks is essential for accurately simulating the mobility of tracked vehicles, particularly in off-road environments. Traditional high-fidelity models provide detailed insights but are computationally expensive, making them unsuitable for real-time applications. This work presents an efficient approach for simulating flexible tracks by dynamically calculating track shape, tension, and contact forces while maintaining computational efficiency.

The track shape is determined based on an ordinary differential equation (ODE) that relates wheel positions, track tension, wheel sinkage, and pressure distribution. Ray casting is used to estimate sinkage along the track, while track forces (including normal, shear, and lateral forces) are computed using established pressure-sinkage and shear stress-displacement relationships. The model updates track shape each time step by estimating entrance and exit angles at wheel contact points and fitting a polynomial to ensure smooth transitions. Contact forces are then distributed to the wheels, enabling accurate simulation of traction and mobility.

This approach efficiently captures key deformation behaviours observed in high-fidelity models while achieving real-time performance by incorporating track tension effects and interaction forces without explicitly solving complex structural deformations. The method is well-suited for vehicle dynamics studies, robotics, and autonomous systems where fast and reliable terrain interaction modelling is required.

ABSTRACT. Existing earthmoving autonomy is largely confined to highly controlled and well-characterized environments due to the complexity of vehicle-terrain interaction dynamics and the partial observability of the terrain resulting from unknown and spatially varying soil conditions. To overcome these restrictions and facilitate development of more robust autonomous earthmoving, a soil-property mapping system is proposed to extend the environmental state.

A GPU accelerated elevation mapping system is extended to incorporate a blind mapping component which traces the movement of the blade through the terrain to displace and erode intersected soil, enabling separately tracking undisturbed and disturbed soil. Each interaction is approximated as a flat blade moving through a locally homogeneous soil, enabling modelling of cutting forces using the fundamental equation of earthmoving (FEE). Building upon our prior work on in situ soil-property estimation, a method is devised to extract approximate geometric parameters of the model given the uneven terrain, and an improved physics infused neural network (PINN) model is developed to predict soil properties and uncertainties of these estimates.

A simulation of a compact track loader (CTL) with a blade attachment is used to collect data to train the PINN model. Post-training, the model is leveraged online by the mapping system to track soil property estimates spatially as separate layers in the map, with updates being performed in a Bayesian manner. Initial experiments show that the system accurately highlights regions requiring higher relative interaction forces, indicating the promise of this approach in enabling soil-aware planning for autonomous terrain shaping.

ABSTRACT. The study focuses on simulating the mobility systems of vehicles on soft terrains, which necessitates the use of physics-based soil models, tire models, and their interactions. A significant challenge lies in obtaining soil model parameters for empirical traction equations and high-fidelity DEM-FEM techniques, mainly due to the variability in soil initial conditions, such as soil bulk density. The objective is to simulate soil behaviors and traction forces related to pull and motion resistance using reduced mathematical models and DEM-FEM techniques. Bekker’s soil and DEM contact soil model parameters were calibrated using soil mechanics testing data on cohesive-frictional artificial soil. Predicted soil rut depth and traction forces from the two modeling techniques were validated against soil bin data from single tire testing using a Yokohama Geolandar radial tire (235/75R15) under two vertical loads (4 kN and 6 kN) and two tire inflation pressures (179 kPa and 241 kPa), with the tire operating in towing mode on dense and loose initial conditions. DEM elasto-plastic soil model and FEM PM-tire model demonstrated better accuracy in predicting soil sinkage compared to the Bekker parameterized soil modeling technique. The Bekker approach provided relatively accurate traction force predictions for free-towing scenarios, particularly in denser soil conditions and with over-inflated tire pressure compared to correctly inflated conditions. The DEM-FEM modeling effectively captured soil behaviors, including compression and bulldozing, in a loose soil state. The comparison of Bekker’s reduced physics and high-fidelity DEM-FEM approaches for predicting soil-to-tire interactions offers opportunities to accelerate the automation of mobility systems on deformable terrains.

ABSTRACT. Lunar exploration, which began in the 1960s, is still ongoing and various space exploration programs are planned. In these space exploration programs, it is important to analyze and predict the trafficability of exploration rovers on the lunar and Martian surfaces, and to develop and operate on these ground conditions. Therefore, in this study, we focused on the real-time analysis in operational phase of exploration rovers. We applied a Long Short-term Memory (LSTM)1) network, an artificial regression neural network architecture used in the field of deep learning, to single-wheel traveling phenomena with the aim of applying it to the real-time prediction in future exploration rover operations. Specifically, numerical simulations were carried out using an extended terramechanics model that takes terrain surface deformation into account, and training and prediction are carried out on several LSTM models. The results show the possibility of real-time prediction of trafficability of a wheel on soft ground due to the characteristics of LSTM networks that can handle time-series data. Furthermore, it is suggested that the proposed method may be applicable to the prediction of precursor phenomena of wheel stuck.

1) Gonzalez, J ; Yu, W. Non-linear system modeling using LSTM neural networks. IFAC-Pap. 2018, 51, 485-489.

ABSTRACT. This abstract summarizes the theoretical approach and method of application used to identify available travel paths, based on the near real time assessment of the interaction between the vehicle and the deformable terrain surface. The framework of this system is based on two avenues of effort that are combined to provide high fidelity mobility corridors that the manned or unmanned vehicle can use to help ensure successful mission completion. The first avenue uses multiple layers of terrain data to determine parameters of map grid elements, including but not limited to: soil type, elevation, soil moisture, roughness, slope, and land cover. These layers are used to establish conditions the vehicle may encounter. The second avenue of effort is a 3DMDB physics-based, deformable soils vehicle model. A large array of operational conditions, are provided to the high fidelity model to produce a predicted mobility and operating speed for each. The resultant information is used to create a fast running model, which is then combined with the first avenue. By combining both avenues, the area of operation can be evaluated to determine the available terrain and the speed at which a vehicle can maneuver through a selected area. The fast running model is continuously updated with available vehicle parameters and sensor data, to assess the capability of the vehicle to continue through the operational area. The real time interaction of the vehicle with the terrain is assessed using established terramechanics, physics-based relationships, updating the prediction for the available routes, and speed which can be achieved. Using a cost based optimization model, a route based on the selectable criteria such as fastest or most efficient is then computed resulting in a recommended path.

ABSTRACT. Understanding the geotechnical properties of lunar soils is essential for predicting their mechanical behavior under varying stress conditions, which is critical for future surface operations such as mobility, construction, mining, and foundation design. However, due to the Moon's reduced gravity, the mechanical behavior of regolith may differ significantly from that observed under Earth conditions. To address this challenge, this study uses Discrete Element Method (DEM) simulations to analyze and calibrate the behavior of lunar regolith simulants under terrestrial gravity to enable more realistic modeling under lunar conditions. Triaxial compression tests were performed on six lunar regolith simulants to determine shear strength parameters, including cohesion, internal friction angle, and stiffness. These laboratory results served as calibration targets for DEM simulations in EDEM software. DEM models were iteratively adjusted to match experimentally the stress-strain responses obtained, accounting for particle size distribution and shape characteristics. A subsequent sensitivity analysis was conducted to identify micromechanical parameters' influence on model accuracy, particle friction, rolling resistance, and contact stiffness. The outcomes provide a validated calibration framework and highlight which parameters are most critical for replicating realistic regolith behavior. This work contributes to developing high-fidelity DEM models for lunar soil and supports their integration into rover mobility studies and surface engineering applications.

ABSTRACT. To prevent excess wear and tear of paved roadways and increased maintenance of winter service vehicles, snowplows typically utilize a bolt on straight hardened metal blade affixed to the 'cutting', or leading edge of the plow. However, when compared to serrated blade leading edges, straight blades leave a compacted frozen surface left adhered to the road requiring further deicing by spreading of sand or road salt. For weather conditions or surfaces when wear of the blade or roadway is not the primary concern, such as in battlefield engineering, serrated plowing edges may be more effective at rapidly creating higher traction roadway surfaces than straight blades. Serrated blades are low cost to manufacture and take only minutes for plow operators to bolt on or remove In March 2025, researchers at the Cold Regions Research & Engineering Laboratory conducted performance testing of an instrumented light US Army tracked vehicle on two surfaces plowed separately with serrated and straight blades. Researchers conducted Drawbar Pull and rolling resistance tests on each surface to measure tractive power of the instrumented vehicle. The effects of using a serrated edge plow blade are seen to be similar to the combined effects of using a straight plow blade and a separate grooming vehicle. Traction increases seen from using serrated edge plow blades for snow and ice removal operations will be shown to be particularly advantageous in tactical domain environments such as plowing over dirt, frozen ground, ice layers, or older high density snow pack.

ABSTRACT. In FY 2026, the Martian Moons eXploration (MMX) is scheduled to be launched. In the MMX mission, the spacecraft will collect a sample from Phobos, a Martian Moon, to bring back to Earth, and the safe landing of the lander is crucial to the success of the mission. However, the MMX mission and other planned landing missions on low-gravity solid bodies have difficulties in evaluating landing performance in advance using ground experiments and simulations due to uncertainties in the characteristics of the target surface layer. In this study, we examined an estimation method for the surface conditions based on the data assimilation analysis using the collision data of an advance landing probe as a precursor before landing on the main lander. Specifically, we attempted to estimate the stiffness and bulk density of the ground for a drop tower experiment using the footpad of MMX lander, by integrating the collision model based on the resistive force theory and the particle filter. In the drop tower experiments, Phobos-simulated regolith was used as the test soil under quasi-vacuum conditions (less than 100 Pa) and microgravity conditions (1/2500 G). The results of the data assimilation analysis of time series of impact reaction force and settlement under various bulk density conditions indicate that the proposed approach could estimate the initial state of the regolith and the change in state after impact to some extent. However, the accuracy of this particle filter-based estimation method depends on the performance of the collision model (physical model), and it is necessary to use the collision model that can adequately represent realistic phenomena.

ABSTRACT. The moisture content of soil is the single most important factor for trafficability of fields. Many sensing technologies are limited to surface moisture such as optical techniques or may be sensitive to variation in texture or salinity. Elastic wave methods are more likely to be sensitive to the mechanical properties of the soil and are directly related to the trafficability of the surface. Acoustic methods are also not impacted by the crowding of the electromagnetic spectrum.

Previous investigators have considered the use of acoustic sensing for moisture and hardpan depth. However, the transfer of this research into applications has been limited. This work is initially focused on characterizing the underlying physics of waves in porous media at a length scale applicable to sandy soils with a narrow pore size distribution. These soils are well suited to this investigation because of potential importance of the slow wave in for moisture detection. The slow longitudinal wave has been used extensively in other areas such as geophysics and membrane characterization. The pore size in sandy soil would not lead to propagation of a slow wave at practical sensor frequencies. However, an evanescent slow wave will impact the effective acoustic impedance of the soil air interface. The energy transfer to the evanescent wave will depend on the portion of the pore structure that is filled with water instead of air.

A model for reflection from a porous interface is formulated for a narrow distribution of pore sizes. Sensitivity of the model to different water content and the effect of frequency on the depth of penetration are considered. Simple experiments with a model soil are used to validate modeling, explore sensor configurations and proposes sensor field deployment.

ABSTRACT. The moisture content of soil is the single most important source of uncertainty for the trafficability of fields. Because of increasing frequency and intensity of rain events, planting or field equipment operations are more likely to occur as soon as possible after a rain events. For harvesting, pest control, planting or other field operations timeliness is critical for the effectiveness of the work and ultimately the yield produced. The trafficability necessary to complete field work is largely a factor of soil moisture from the surface down to 15 to 30 cm; however, large temporal and spatial variations complicate sensing and deploying machinery in the areas that are sufficiently dry to support fieldwork.

This work focuses on high sand content soils: sandy loam, loamy sand and sand. Under controlled conditions L-band Microwave Radiometry was used to measure the moisture content in a non-contact configuration. Initial testing focused on the effect of interference from cellular, emergency and other radio sources of radio frequency interference. A second test was conducted in proximity to general aviation operations. Initial concerns regarding noise from most sources was found to be less of a problem than anticipated.

However, at the frequency employed (1.4 GHz), a poor correlation with time domain reflectometry at 3 cm depth was observed, suggesting that the sensing depth of the microwave sensor is deeper than the TDR probe. Higher microwave frequencies may need to be employed to match the critical depth for sandy soils, A number of alternative applications exist for these moisture sensors which may be of value in other applications or surfaces with higher clay content or higher salinity.

ABSTRACT. Delivering resilient, off-road autonomy for unmanned ground vehicles (UGVs) will not come from incremental advances in perception, planning, or mobility alone. The decisive enabler is a robust software operating environment—a foundational architecture that integrates autonomy stacks, payloads, and edge compute into a coherent and resilient system-of-systems. Today, much of the DOD’s ground robotics development remains platform-centric, with limited ability to reuse or scale autonomy services across missions. Without a modular, standards-aligned software backbone, breakthroughs in autonomy cannot reliably transition to the field. Foundational software architecture must be treated as the cornerstone of UGV development. Open, containerized, and portable environments enable rapid integration of autonomy algorithms, heterogeneous payloads, and resilient communications, while ensuring certification pathways for cyber and safety. The lessons learned are informed by Parry Labs’ prior work modernizing the MQ-1C Gray Eagle fleet, where software-first approaches demonstrated how containerization, abstraction, and modular integration could transform a legacy platform into a flexible, multi-mission system. Drawing on these insights and previous experience with the Ground Vehicle Systems Center and Marine Corps Warfighting Lab's previous and ongoing robotics and UGV efforts, we will outline both the risks of not adopting a software-first paradigm and the operational advantages gained when architectures are designed for reuse, adaptability, and contested environments. Ultimately, a foundational software operating environment is the enabler of resilient autonomy, accelerated capability delivery, and true multi-mission relevance for the Army’s future ground vehicle fleet.

ABSTRACT. We present results from vehicle testing conducted on an Arctic ice road in February 2025 and monitored with a fiber optic distributed acoustic sensing (DAS) system. The DAS sensing element, a 1-km telecommunications-grade fiber-optic cable, was deployed in three configurations: surface-laid on snow, surface-laid on ice, and frozen into ice. We utilized an amplitude-only DAS system that recorded strain-rate time series with a channel spacing of 5 m. Detection metrics as a function of deployment configuration were derived against multiple vehicles transiting at variable speeds. Detection statistics are shown to vary with deployment conditions and target characteristics, suggesting that sensor performance can be optimized by accounting for environment and source type. Additionally, we used the fiber-optic sensor and vehicle excitations to interrogate the propagation medium. We present evidence that observed differences in frequency-dependent transmission loss can discriminate DAS cables embedded in ice and snow. This work resulted from an international collaboration enabled by the International Cooperative Engagement Program for Polar Research (ICE-PPR).

ABSTRACT. Winter surfaces are characterized by low strength, which can result in deep vehicle sinkage and high vehicle motion resistances. If the vehicle’s sinkage is greater than the vehicle’s ground clearance, snowplowing forces occur, further increasing the vehicles motion resistance. The increased motion resistance can lead to vehicle immobility if the vehicle’s available traction is exceeded. In this work we investigate the effect of snow subduction on vehicle mobility. When snowplowing occurs a portion of the plowed snow can be subducted under the vehicle’s belly. The subducted snow is compressed resulting in both an increase in motion resistance and a normal load reduction under the vehicle’s track or tire. We developed an algorithm which includes both snow subduction effects on the vehicle’s mobility. In the study we found that the snow subducted under the vehicle has a substantial effect on the vehicle’s mobility. We simulated a 103kPa ground pressure vehicle with 0.1m snow subduction zone. When snow subduction was included in the analysis, we found that vehicle immobility occurred at a snow depth of 0.59m as compared to a depth of 0.7m when subduction was not included. The large variation in snow depth at vehicle immobility, underlines the importance of including snow subduction in vehicle mobility and route planning analyses. In future work we will improve the accuracy of the algorithm by incorporating snow stratigraphy. The current algorithm uses bulk density to determine snow strength, in future analyses we will incorporate snow layer density and grain size to improve the algorithms accuracy.

ABSTRACT. Ride quality is an important metric for assessing the comfort and safety of vehicle passengers as a vehicle traverses rough off-road terrain. One common approach for assessing ride quality is the Absorbed Power metric, which sets a power threshold below which a crew member can effectively perform their tasks. The roughness of the terrain itself is characterized by the root mean square (RMS) value of the terrain profile as per TOP-1-1-014. Using an accelerometer to measure the acceleration experienced by the passengers, a series of tests can be conducted in which the vehicle and traverses a given course at constant speeds, increasing the speed with each traversal. By calculating the absorbed power at each speed, a maximum allowable speed can be determined for a given course. Because it is not always possible to take a vehicle into the field to test it on such a course, vibration testing using a four-post shaker to emulate the terrain profile can be used as an alternate test to assess the ride quality. In this work, we perform vibration testing on a light utility vehicle in an effort to assess ride quality over an asymmetric 3cm RMS terrain. Additional physical testing – including a tilt test to determine center of gravity height – were performed to tune a digital twin of the vehicle. The same suite of tests was performed on the digital twin in a virtual environment and compared to the physical test results. The virtual test of the vehicle traversing the 3cm RMS course that the vibration test emulates was also performed on the digital twin to compare vibration testing results to that of a fully modelled traversal of the terrain.

ABSTRACT. The abstraction of the world in a virtual space known as a map is crucial to enable motion planning. Autonomous mobility over uneven terrains requires the model of terrain topography to account for surface profile induced factors that affect the feasibility of planned trajectories. Point cloud-based environment modeling has been particularly effective due to their ability to produce depth information. However, when deployed in UGVs, they encapsulate both traversable terrain and positive obstacles which must then be distinguished. This work introduces a novel and fast point cloud segmentation algorithm called the inverse ray tracing algorithm to classify traversable and non-traversable points.

The segmented point cloud is then collated together to create a comprehensive map. To this end, we present a Bayesian method using inverse sensor models that generate a 3D voxel grid of the terrain’s topography and a second method using convex hulls around classified terrain points, sequentially appending points to generate the terrain topography map. The proposed results are suitable for UGVs operating in GNSS-denied environments. Our approach integrates sensor fusion from a visual-inertial sensor, wheel encoders, and an inertial measurement unit with an extended Kalman filter to maintain precise odometry. The effectiveness of the proposed algorithms has been experimentally validated on a Terramechanics rig and compared with point cloud obtained from a 3D scanner, which resulted in a chamfer distance of approximately 1 mm, demonstrating high accuracy. Additionally, the navigation stack has been deployed on the Clearpath Husky platform to generate real-time terrain topography and obstacle maps, proving the practical applicability and robustness of our proposed algorithms.

ABSTRACT. Understanding interactions of wheels and tracks with granular media under variable loading conditions, including longitudinal and side slip, is critical for prediction of mobility for wheeled and tracked vehicles in off-road environments. The discrete element method (DEM) is routinely used for modeling interaction of soil with track and tires, but the method’s accuracy needs to be better established. In this work, two DEM models from the Generic EDEM Material Model (GEMM) database from Altair®'s EDEM™ software package, which were identified (Jelinek et el. 2025) as the best match to physical experiments, are used to calculate tractive performance of a rigid wheel in sand under braked, towed, and powered conditions with side slip. The simulations follow the experiments by Shinone et al., 2010, examining a 165/60R13 wheel with constant circumferential velocity of 97.6 mm/s and vertical contact load of 980 N operating in powered conditions under forward slips in the range of -5.9% to 54.8%. The steady-state tractive forces on the wheel in sand were evaluated under the same forward slip conditions as in the experiment, except with added side slip of 3, 6, and 12 degrees. The side-slip effects on net traction, gross traction, and sinkage of the wheel in sand are evaluated by comparison with forward-slip-only results. Lateral forces and overturning moments were then calculated and compared with relevant relationships derived from physical measurement and other simulation methods. The results indicate a good match to experimental trends, encouraging further use and calibration of the DEM soil-wheel interaction models.

ABSTRACT. Proper Discrete Element Method (DEM) calibration is essential for guiding reliable engineering decisions on the optimal design of tools and wheels interacting with soil. Optimally designed tools and wheels save time, reduce costs and lower emissions. The paper highlights and validates two overlooked contributions to the DEM calibration literature: the relationship between physical and DEM relative density, and the scale-invariant effect of friction on critical state mechanics. These contributions significantly reduce calibration iterations and enhance the accuracy of soil simulations. The research includes direct shear tests on Michigan 2NS sand and blade mixing experiments to validate the DEM calibration. The results show that mechanistic, non-iterative approach to DEM calibration can yield an accurate DEM parameter set that can be further refined to meet specific application needs.

ABSTRACT. The Army currently has limited capabilities to calculate training impacts for multi-day-vehicle exercises. The ARNG has deployed a program of record for eXportable Combat Training Capability (XCTC) which utilizes a tracking system similar to those used for fleet management. The use of spatial and temporal data from these training events can then be linked to vehicle severity models developed by CERL, CRREL, and other agencies to quantify actual impacts to ground, air, and water. CERL is utilizing data from historic exercises to develop a machine learning capability that could then project out impacts to the terrain from existing and future vehicle platforms. This presentation will discuss lessons learned and accuracy of the model estimations for sinkage and disturbance.

ABSTRACT. In the realm of civil protection at countries with security concerns like Colombia, armored vehicles play a crucial role in ensuring the safety and security of persons. These vehicles undergo significant modifications, including increased mass to enhance their armor features, which in turn alters their dynamic behavior. This study explores the feasibility of using an off-road vehicle to simulate the dynamic behavior of an armored vehicle, thereby providing a cost-effective platform for the development, and testing of non-sprung components. The project utilized a modified off-road vehicle with an initial mass of 2,500 kg. By incrementally adding ballast, the vehicle's mass was increased to 3,500 kg, simulating the weight of an armored vehicle. Dynamic tests were conducted on three types of terrains (gravel, mud, and asphalt), to evaluate the vehicle's performance under different conditions. Key parameters such as suspension travel, damping characteristics, and ride comfort were measured and analyzed. Preliminary results indicate that the modified off-road vehicle successfully replicates the dynamic behavior of an armored vehicle with a reliable degree of accuracy. Components, like the shock absorber increases their load and could harmful the vehicle dynamics for this type of vehicles. With this methodology it could be possible to analyze too the energy absorption and ride comfort metrics for these types of vehicles. This research demonstrates the potential of using older off-road vehicles as a viable platform for simulating the dynamic behavior of armored vehicles. The findings provide valuable insights for the development of non-sprung components, ultimately contributing to the enhancement of armored vehicle performance in civil protection applications.