Online Elasticity Estimation and Material Sorting Using Standard Robot Grippers (2401.08298v3)

Abstract: We experimentally evaluated the accuracy with which material properties can be estimated through object compression by two standard parallel jaw grippers and a force/torque sensor mounted at the robot wrist, with a professional biaxial compression device used as reference. Gripper effort versus position curves were obtained and transformed into stress/strain curves. The modulus of elasticity was estimated at different strain points and the effect of multiple compression cycles (precycling), compression speed, and the gripper surface area on estimation was studied. Viscoelasticity was estimated using the energy absorbed in a compression/decompression cycle, the Kelvin-Voigt, and Hunt-Crossley models. We found that: (1) slower compression speeds improved elasticity estimation, while precycling or surface area did not; (2) the robot grippers, even after calibration, were found to have a limited capability of delivering accurate estimates of absolute values of Young's modulus and viscoelasticity; (3) relative ordering of material characteristics was largely consistent across different grippers; (4) despite the nonlinear characteristics of deformable objects, fitting linear stress/strain approximations led to more stable results than local estimates of Young's modulus; (5) the Hunt-Crossley model worked best to estimate viscoelasticity, from a single object compression. A two-dimensional space formed by elasticity and viscoelasticity estimates obtained from a single grasp is advantageous for the discrimination of the object material properties. We demonstrated the applicability of our findings in a mock single stream recycling scenario, where plastic, paper, and metal objects were correctly separated from a single grasp, even when compressed at different locations on the object. The data and code are publicly available.

Evaluating Online Elasticity Estimation of Soft Objects Using Standard Robot Grippers

The paper investigates the feasibility of using standard robot grippers for online elasticity estimation of soft objects, a task normally outside the designed function of these robotic components. Elasticity estimation is essential for applications such as object recognition and sorting in manufacturing environments where tactile information is more reliable than visual data alone. This research systematically evaluates the potential of using sophisticated biaxial compression devices as control instruments against standard parallel jaw grippers with mounted force/torque sensors to gauge material characteristics of deformable objects.

Experimental Setup and Methodology

The authors utilize two main gripper types: the OnRobot RG6 and Robotiq 2F-85, alongside a force/torque sensor to assess their performance in capturing mechanical responses from soft objects. An assortment of deformable objects, including ordinary cubes and polyurethane foams, were subjected to multiple experimental setups. Key parameters manipulated during experimentation included speed and repetition of compression cycles, as well as gripper surface area.

During the evaluation, stress/strain curves were constructed, challenging the devices to measure parameters such as Young's modulus at various strain points and viscoelasticity through energy absorption analysis. The rigorous methodology also involved using a professional setup to establish control measurements of elasticity and viscoelasticity to accurately compare the grippers' performance.

Key Findings and Analysis

One significant finding is the relative success rather than the absolute accuracy of the grippers in estimating material elasticity. While robot grippers showed limitations in delivering accurate specific estimates of Young's modulus, they excelled in consistently ordering materials by elasticity reliably across different device setups. This was notably achieved at lower compression speeds, whereas the surface area exerted no substantial beneficial impact on estimate precision.

In terms of viscoelasticity, the Hunt-Crossley model provided a satisfactory non-linear characterization of viscoelastic properties over other modelings, like the Kelvin-Voigt model. The dual analysis of elasticity and viscoelasticity furnished a two-dimensional space advantageous for discriminating between various deformable objects. A noteworthy contribution is the use of a publicly available dataset and processing code, potentially benefiting others in the research community facing similar challenges in material property estimation.

Practical Implications and Future Directions

Although robot grippers cannot directly measure the Young's modulus with high precision, their use in relative material ordering holds substantial practical utility, especially in sorting or handling tasks in production lines. The success of this approach points toward future enhancements where integrating tactile sensors at gripper fingertips could further refine force feedback, enhancing measurement accuracies.

The paper indicates a promising direction for applying federated robotic sensing in applications requiring tactile exploration, differentiating materials based on their deformations. Future research could expand on this paper by leveraging enhanced tactile sensing technologies and exploring machine learning approaches to improve the fidelity of material characteristic estimation, potentially making more robust and flexible robotic solutions available to industry. In summary, this paper frames a measured and systematic exploration into a niche application of robotic grippers, providing both critical insights and a foundation for subsequent technological advancements in haptic-based material analysis.