TOWARDS THE REALIZATION OF AN ARTIFICIAL TACTILE SYSTEM - FINE-FORM DISCRIMINATION BY A TENSORIAL TACTILE SENSOR ARRAY AND NEURAL INVERSIONALGORITHMS

This paper describes techniques and methodologies so far developed to investigate object fine-form discrimination by means of artificial tac tile sensors. Sensor arrays, selectively sensitive to stress-tenser co mponents and based on piezoelectric polymer technology, have been real ized. Sensor output data are used to solve inverse elastic contact pro blems, by means of neural networks suitably trained to learn regulariz ed inverse maps. Two possible neural network designs are considered: o ne is based on the multi-layer perceptron trained with the standard ba ckpropagation algorithm, and the other is based on the use of radial b asis functions. In both cases, reconstruction of object shapes is demo nstrated to be effective and robust with both simulated and real data.