Shape self-sensing with mutual inductance sensor array

Real-time sensing of shape is an important tool for many intelligent machines, particularly in soft robotics. Mutual induction data from an array of sensors shows great promise as an accurate tool for shape sensing. In this paper, we show how inductive array data can be used for shape imaging and topographic shape tracking. The idea has been extended to many geometrical settings showing a versatile tool for shape sensing. The sensors are arranged around a circular array allowing reconstruction of the deformation from circular shape to generic polygon shape including elliptic shape. A linear array shows the sensing of tension force and various deformation of lines. Finally, the sensor array is used on a surface allowing reconstruction of both shear force and the normal force to the surface. A suitable method of calculation of the mutual inductance between two coils has been implemented and a range of methods including inversion algorithms, calibration methods, and a machine learning tool show the application of the new shape sensor system.