Fig. 6: E-Antenna (millimeter-scale) enabled mobile robots with texture perception capability.

A Functionlize cleaning robots with texture perception can make them smartly adopt the adaptive cleaning strategy on different surfaces (i): a common scenario in our residential spaces with various types of floors. ii): If the existing sweeping robots have the ability of texture perception, they can choose the sweeping strategy autonomously). B Integrate E-Antenna (1 mm in diameter and 8 mm in length) underside of a mobile robot to functionalize it with ground texture perception capability. C Three common floors in residential scenes, i.e., carpet, ceramic floor, and wooden floor, were used for testing. D Samples of tactile signals on these three floors show the E-Antenna can capture the fine difference between these floors. E Visualization of the outputs of the trained network of ATPM with t-SNE in response to these three floors, where blue dots represents the ceramic tiles, purple dots prepresents carpet, and yellow dots represents wooden. F Classification confusion matrix of the trained network for these three floors, where most cases achieve 100% perception accuracy, indicating that the E-Antenna can enhance the mobile robot with reliable ground perception ability.