A somatosensory system for a robot

Project: Research

Project Details


In this project we are investigating the possibilities of creating a somatosensory system for the humanoid robot Epi. This is an experimental and technical endeavour which concerns whether and how capacitive sensors can be used as to detect different types of touch, and the treatment of the signal thereby created, including cognitive modelling. This is an investigation of physical cognition in robots, which has implications for social cognition and social interaction between humans and robots.

The capacitive sensors will be created using conductive ink and connected to an Arduino card from Bare Conductive, which has twelve touch sensitive electrodes. The robot head shell is created from plastic in a 3D printer, where different materials could possibly be used and the form of the head be adjusted. We have investigated how well a change in capacitance can be detected through the robot head shell, with a thickness of 1-2 mm, and can observe a clearly distinguishable signal on the inside of the head as it is being touched on the outside.

An important question is that of how the sensor areas should be placed as to best detect different kinds of touch. This involves developing a theory of what kinds of touch is relevant and of interest to detect, and adjusting the form and placement of sensory areas accordingly.

We have decided to use twelve different areas, each corresponding to an electrode on the Arduino board, covering larger areas of the inside of the shell, but leaving 2-6 cm between the different areas unpainted, as to clearly differentiate between them.
This would primarily enable detection of which area has been touched and could also be used for detecting motion between different areas. The frequency of touch could also serve as an indicator as to what patterns of touch are being applied, as in for example patting the head.

Questions regarding more complicated forms of touch include the resolution of the signal and of how the capacitance changes with for example a larger area of touch or with stroking on one area.
Effective start/end date2020/01/012020/06/30