Figure 1. Piezoresistive sensors mounted on the fingers of a model robotic hand. (Credit: Bao research group, Stanford University.)
Figure 1. Piezoresistive sensors mounted on the fingers of a model robotic hand. (Credit: Bao research group, Stanford University.)

A skin-like polymeric material is using carbon nanotubes (CNTs) to bring a sense of touch to robotic and prosthetic devices. Developed by researchers at Stanford University and Xerox Palo Alto Research Center, the flexible, polymeric skin or ‘digital tactile system’ (DiTact) incorporates CNT pressure sensors and flexible organic printed circuits to mimic human response [Tee et al., Science 350 (2015) 313].

‘‘We wanted to make a sensor skin that communicates in the same way as the body,’’ explains research student Alex Chortos, one of the lead authors of the work. ‘‘The goal is to make skin for prosthetics that can feel touch in a natural way and communicate that information to the person wearing the prosthetic device.’’

In the body, receptors in the skin relay sensing information directly to the brain in a series of voltage pulses rather like Morse code. Artificial devices employ tactile sensing to improve the control of neuroprosthetics and relieve phantom limb pain. But, to date, prosthetic skin devices have had to use a computer or microprocessor to turn the output from sensors into a signal compatible with neurons.

The new approach, by contrast, combines these operations in a single system of piezoresistive pressure sensors embedded in a flexible circuit layer. The sensors are made from a CNT composite dispersed in a flexible polyurethane plastic and molded into pyramidal structures. The pyramidal shape is crucial because it allows the pressure range of the sensor to be tuned to that of skin.

The operation of the device is simple: pressure on the sensor squeezes the CNTs closer together, allowing an electric current to flow. Changes in pressure are translated directly into digital signals, the frequency of which varies with the intensity ofthe stimulation, mimicking the behavior of tactile receptors in the skin.

When put into actual devices such as a prosthetic hand or wearable glove (Fig. 1), as the team led by Zhenan Bao demonstrates, the sensors can detect the difference between a soft touch and a firm handshake.

The signals generated by the sensors can be relayed externally using an inkjet-printed flexible organic circuit layer employing stretchable silver nanowire conductors. The digital signal from the system can even be used to stimulate neurons in mouse brain tissue directly, both electrically and optically.

Usually prosthetic devices that interface with the brain do so via electrical stimulation. But in an intriguing new approach, the team tried out a technique known as ‘optogenetic stimulation’, where a digital signal is used to modulate an LED source that produces neural firing via light-activated ion channels.

The new system has a number of potential advantages, explains Chortos. ‘‘It is a simple system that communicates information in a way that brain cells can understand. It is made of plastic materials, so it could be made relatively cheaply. And the sensor is not susceptible to noise, so that it can communicate information over long distances and still be accurate.’’

Just like real skin, DiTact is also very efficient, he adds. The researchers’ calculate that their system could operate for 500 days on the equivalent of a single iPhone battery charge.

‘‘The work is quite impressive,’’ says John A. Rogers of the University of Illinois at Urbana-Champaign, ‘‘simply in terms of the technical sophistication of the system level demonstrators in which they’ve been able to combine organic electronic oscillators with soft pressure sensors.’’

The team is now working to improve the system by reducing the sensor size and enhancing its sensitivity so that more subtle moving stimuli (like stroking) can be distinguished from simple static pressure.

This article was originally published in Nano Today (2015), doi:10.1016/j.nantod.2015.10.004