Low‐Strain Flexible Ink‐Based Sensors Enable Hyperelastic Inflation Perception Via Geometric Patterning for Balloon‐Type Robots

Integrating proprioceptive capabilities in balloon‐type robots holds immense potential to enable safe interactions between humans and machines. These robots, made of elastic or hyperelastic materials, undergo substantial deformations when subjected to changes in pressure. However, simply estimating balloon deformations based on pressure changes is inadequate due to the nonlinear relationship between pressure changes and balloon deformations. Therefore, embedding stretchable sensors on the balloon's surface is a more effective approach. This deformation stretches the surface of the robots and activates the embedded flexible sensors. Nonetheless, mismatched stretching characteristics between the sensor materials and the balloon surface can result in sensor damage or failure. To address this challenge, a novel design strategy to enhance the contact perception capability of soft balloon systems by optimizing the geometry and assembly layout of flexible sensors. Numerical calculations to assess the strain variation of the Ω shape are performed along the intersection arrangement compared to the typical strip shape along the radial arrangement for the tensile rate comparison to determine the desired size and arrangement. The fabrication of conductive ink and calibration on flexible sensing is introduced to fabricate sensors with customized shapes. The light‐curing transfer‐printing strategy is explored to precisely attach the sensors to specific intersections of the balloon surface. Finally, flexible sensors of shape with a stretch rate of ≈127 are successfully employed to estimate the surface expansion rate of over 300 on a balloon‐type robot. The integrated system exhibits a near‐linear sensitivity below 318 kPa with change rates of 0.7525 mm/kPa in circumference, 13.3826 /kPa in cross‐sectional area, and 0.1484 k/kPa in resistance. The linearity error rate after 400 kPa is around 9.7. Furthermore, the ablation experiment showed circumference and area errors of less than 5 during the inflation procedure. Pneumatic cycle test experiments were introduced to verify stability and repeatability. Furthermore, in multiple contact perception experiments under pressure switching modes, the system exhibited a substantial rate of change, reaching around 146 and 128. These advancements provide a new perspective for future designs of flexible sensors and contribute to developing safe and user‐friendly interactive robots.