Tactile soft-sparse mean fluid-flow imaging with a robotic whisker array

An array of whiskers is critical to many mammals to survive in their environment. However, current engineered systems generally employ vision, radar or sonar to explore the surroundings, not having sufficiently benefited from tactile perception. Inspired by the whisking animals, we present here a novel tomography-based tactile fluid-flow imaging technique for the reconstruction of surroundings with an artificial whisker array. The moment sensed at the whisker base is the weighted integral of the drag force per length, which is proportional to the relative velocity squared on a whisker segment. We demonstrate that the 2D cross-sectional mean fluid-flow velocity-field can be successfully mapped out by collecting moment measurements at different angular positions with the whisker array. We use a regularized version of the FOCal underdetermined system solver algorithm with a smoothness constraint to obtain soft-sparse static estimates of the 2D cross-sectional velocity-squared distribution. This new proposed approach has the strong potential to be an alternative environmental sensing technology, particularly in dark or murky environments.