Tactile sensors have been a source of inspiration for developing novel (and old) concepts based on nanotechnologies for smart systems integration. The perception of human senses (heat, textures, softness, humidity) can highly benefit from biomimetic strategies, where physical sensors can provide information on the surroundings. In this talk, magnetic–based soft sensors are integrated as a biomimetic, skin-inspired magnetic tactile sensor. Examples of devices comprising a 4´4 matrix of magnetic tunnel junction TMR sensors are integrated below Ecoflex skin impregnated with 5 μm NdFeB hard magnetic particles. The challenges in integrating of TMR sensors for 2D detection using 1D sensor chips, and also the integration of full Wheatstone bridges for 2D detection will be discussed.

The tactile capability for texture assessment and the device robustness for harsh tasks were evaluated using a 3D-stage for control of an indenting tip mounted at a ATI nano 17 force sensor. The force characteristics in 2D was used to train an artificial neural network models that could predict the applied force using 31260 data points, where the tangential component was [0,2.75] N and the normal component was [0,17] N. For a practical integration, the mechanical stability and plasticity limit of the skin interface was evaluated, with an excellent record of 98 N. This offers a realm of opportunities for practical integration of this magnetic-based tactile sensors in unsupervised environments.