Floor Friction Assessment from Robot Tire-Surface Interaction

Kulbhushan Komarraju; Nathaniel Pearson; Yu Gu; Jason Gross

cs.ROOct 24, 2026

Floor Friction Assessment from Robot Tire-Surface Interaction

Authors: Kulbhushan Komarraju, Nathaniel Pearson, Yu Gu, Jason Gross

Organizations: Department of Mechanical, Materials and Aerospace Engineering, West Virginia University · Lane Department of Computer Science and Electrical Engineering, West Virginia University

Abstract

Floor traction influences wheel-slip behavior, dead-reckoning accuracy, and the navigation performance of wheeled mobile robots that rely on wheel odometry. As these robots become increasingly common in GNSS-denied indoor environments, including automated manufacturing and retail facilities, their onboard measurements could also be used to characterize floor conditions and improve localization. In this study, a skid-steer robot was driven over five floor surfaces with distinct traction characteristics using commanded velocity profiles designed to induce wheel slip through repeated acceleration and direction reversal. Differences between the body velocity estimated from inertial sensing and the velocity obtained from wheel odometry were evaluated using the integrated relative-velocity error, phase-derived wheel-to-body response delay, and velocity-reversal response delay. Welch power spectral density analysis was additionally applied to characterize the frequency distribution of the velocity mismatch during positive and negative acceleration. Band-integrated spectral power, spectral centroid, and spectral decay rate were extracted from each recording and evaluated using linear discriminant analysis and multiclass classification. The response-delay measurements produced consistent surface-dependent trends, with the longest delays observed on soap and the shortest delays observed on carpet. The spectral features also provided clear separation among the tested surfaces, achieving a balanced classification accuracy of 97.1% under recording-level cross-validation. These results demonstrate that both the magnitude and frequency domain characteristics of wheel-to-body velocity mismatch contain useful information about tire-surface interactions. Such information could support surface recognition and the development of slip-aware wheel-odometry weighting within robot state-estimation frameworks.

Figures & tables

Fig. 1 : Skid-steer robot undergoing experiments in the motion capture facility. In this example experiment, the driving surface is covered with a soapy mixture.

Fig. 2 : Schematic of the factor graph estimation being used for robot state estimation. Blue nodes represent the state variables (pose, velocity, and IMU biases) at each timestep. Red factors denote IMU preintegration constraints between consecutive states, while green factors represent Motion Capture (MoCap) pose and velocity measurement updates.

Fig. 3 : Commanded velocity profile

Panel of Figure 4 — Fig. 4 : Velocity comparison and relative velocity profile for Concrete surface.

Panel of Figure 5 — Fig. 5 : Velocity comparison and relative velocity profile for Vinyl surface.

Panel of Figure 6 — Fig. 6 : Velocity comparison and relative velocity profile for Melamine Board surface.

Panel of Figure 7 — Fig. 7 : Velocity comparison and relative velocity profile for Soap surface.

Panel of Figure 8 — Fig. 8 : Velocity comparison and relative velocity profile for Carpet surface.

Fig. 9 : Normalized spectral-power distribution of the relative velocity during negative and positive FG slopes. Each segment was normalized by its total spectral power before the frequency-wise median was calculated. The heatmap therefore represents the distribution of mismatch power across frequency rather than its absolute magnitude. White regions correspond to frequencies excluded because of segment-resolution or sampling limitations.

Panel of Figure 10 — Fig. 10 : Surface discrimination using run-level spectral features. Each point in (a) represents one complete recording. The LDA projection is a supervised visualization of the spectral-feature space. In (b), the matrix entries show the number of held-out recordings assigned to each surface, while the shading represents the row-normalized classification fraction.

Fig. 11 : Median AUC for different surface relative velocity profiles for up and down segments.

Surface	Median $\pm$ Std.
Board	$0.420\pm 0.080$
Carpet	$0.845\pm 0.065$
Concrete	$0.390\pm 0.033$
Soap	$0.150\pm 0.080$
Vinyl	$0.500\pm 0.043$

TABLE I : Median static CoF values across surfaces measured by a drag-sled tribometer.

Fig. 12 : AUC vs SCoF; Horizontal error bars indicate the interquartile range (IQR) of the estimated AUC values, while vertical error bars indicate the standard deviation of the tribometer SCoF measurements.

Panel of Figure 14 — Fig. 13 : Wheel-to-body velocity response delays measured for the five surfaces. Each point represents one complete recording. The boxes show the interquartile range, the center lines represent the medians, and the whiskers extend to 1.5 times the interquartile range. A positive delay indicates that the ground-referenced body velocity responded after the wheel-odometry velocity.

Fig. 14 : Relationship between the static coefficient of friction and (A) the phase-derived wheel-to-body response delay and (B) the velocity-reversal response delay. Each point represents the median delay across complete recordings for one surface. Horizontal error bars indicate the interquartile range of the response delay, while vertical error bars indicate the standard deviation of the measured static coefficient of friction. Dotted lines represent unweighted first-order linear fits.

Mobile Robotics

Mobile robotics studies how ground, aerial, and aquatic robots perceive, navigate, and act in changing real-world environments.

See all 138 papers

SlipSense: Multimodal Sensing for Online Slip Detection in Legged Robots
Jun 23, 2026Iris Szu-Yao Liu, Chien Chern Cheah, Meng Yee Michael ChuahTactile Sensing Robotic Tactile Sensing
School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore · Institute for Infocomm Research, Agency for Science Technology and Research, Singapore
From Impact to Insight: Dynamics-Aware Proprioceptive Terrain Sensing on Granular Media
Apr 2, 2026Yifeng Zhang, Yue Wu, Jake Futterman +5Terrain-Aware Robot Locomotion Granular Terrain Locomotion
Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA 90089, USA
Learning Wheel Motion Uncertainty for Degeneracy-Aware LiDAR-Inertial Odometry on Quadrupedal Wheel-Legged Robots
Oct 19, 2026Hairuo Wei, Yunfan Ren, Renda Yang +2LiDAR Odometry Adaptive Lidar-Inertial Odometry
The Chinese University of Hong Kong, Hong Kong SAR, China. · Tongji University, Shanghai, China. · SEER Robotics.

Jun 23, 2026J/K move · Enter open · S save

SlipSense: Multimodal Sensing for Online Slip Detection in Legged Robots

Iris Szu-Yao Liu, Chien Chern Cheah, Meng Yee Michael Chuah

School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore · Institute for Infocomm Research, Agency for Science Technology and Research, Singapore

Tactile Sensing Robotic Tactile Sensing Quadruped Locomotion Multimodal Sensor Fusion

Open paper AskPDF arXiv

Legged robots rely on accurate ground interaction awareness to traverse variable terrains, such as slippery surfaces. Existing slip detection methods often rely on kinematics and proprioception, which lack the sensitivity to detect early-stage slips that occur prior to catastrophic instability. Thus, this paper presents SlipSense, a novel framework for online force-based slip detection using a custom lightweight sensorized foot for quadrupeds to detect slip. The framework integrates a multimodal sensor design with a LSTM-based model to infer ground reaction forces and detect slip-indicative anomalies during locomotion. The proposed framework is deployed on a Unitree Go1 quadruped to demonstrate blind online slip detection over a slippery terrain. Our method detects early-stage slips down to an average displacement of 24.1 +/-6.4mm with an overall accuracy of 85.9%. This represents a 3.3-fold finer detection resolution and a 24% relative accuracy improvement over a standard kinematic baseline that uses foot velocity inferred through state estimation. The work in this paper serves as a foundation for force-aware gait adaptation in legged robotic locomotion, allowing future controllers to estimate terrain friction and adjust constraints, thus improving the overall stability of the system.

Abstract

Mobile Robotics

Explore similar work

SlipSense: Multimodal Sensing for Online Slip Detection in Legged Robots

From Impact to Insight: Dynamics-Aware Proprioceptive Terrain Sensing on Granular Media

Learning Wheel Motion Uncertainty for Degeneracy-Aware LiDAR-Inertial Odometry on Quadrupedal Wheel-Legged Robots