Maximizing Robotic Limb Rigidity and Strain Sensing Capabilities Through Localized Kevlar Fiber Reinforcement
摘要
Strain can provide important sensory information for locomotor control in both robots and insects. Strain measurements in limbs can indicate forces that resist movements generated by actuators or muscle. Key time points in stepping can be monitored through changes in strain and inform the neuromuscular or robotic control system of the step cycle phase, allowing for modifications during unplanned events. For both systems, there is a trade-off between the limb rigidity and thus accuracy of limb positioning, and the magnitude (i.e., sensitivity) of strain measurements. Rigid limb segments enable precise end effector (e.g., foot) placement but reduce the strain magnitude when force is resisted; compliant limb segments increase the strain magnitude, but extreme compliance can potentially decrease precision in determination of limb position and movement. Robotic limb segments typically have homogeneous material properties, requiring the choice between rigidity and compliance. In contrast, insect limbs have spatial gradients of material properties. We show the benefits of localized Kevlar fiber reinforcement for strain sensing in robotic limb segments. While full-segment Kevlar reinforcement solely increases rigidity, the most effective reinforcement format for increasing both rigidity and strain magnitude is partial fiber supplementation encompassing the majority of the limb segment but not the area where strain is monitored. We propose that future robotic additive manufacturing should incorporate material heterogeneity to optimize rigidity and strain sensing capabilities.