Flexible strain sensor enabled by carbon nanofibers can ‘read lips’

Innovative fabrication techniques with composite materials have enabled wearable flexible strain sensors to monitor the tiny vibrations of human skin. They are discreet and unobtrusive. High sensitivity is key to a high-quality sensor. Unfortunately, due to limitations in structure or conductivity, it can be difficult to get both.

Tsinghua University researchers have recently developed a flexible strain detector design using a membrane of carbon nanofibers (CNF), which is made up of parallel and randomly aligned carbon fibers. This allows for high sensitivity as well as wide range of detection.

The study was published in Nano Research.

Flexible strain sensors are useful in activity and health monitoring, smart textiles, and human-machine interaction. A flexible strain sensor was developed by the Tsinghua University research group for a lip language recognition system. This can be used to help people with damaged vocal chords communicate daily.

Peng Bi, first author at Tsinghua University said that “The lip language recognition system can quickly translate sentences for people suffering from damaged vocal cords.” This significantly reduces the barriers to everyday communication.

Flexible sensors are required to capture information from facial movements, as well as distinguish subtler changes. Bi said that the only way to satisfy this requirement is to create a flexible strain sensing device with both high sensitivity as well as a broad range of strain detection.

Flexible sensors, unlike conventional sensors made out of bulky, rigid metal, can conform to human skin and move without discomfort. These sensors are usually made of elastic polymers and conductive materials such as graphene or carbon nanotubes. They can be integrated into clothing or attached directly to the skin.

Up to now, most reported wearable strain sensors showed either large workable strain range or high sensitivity—but not both. A strain sensor that has a large range of strain detection may be flexible or stretchable by more than 400%, depending on the material. A sensor with a wide strain detection range will usually have a low gauge factor value. This is an indicator of its sensitivity and a limit to the ability to detect vibrations below the skin.

Both of these advantages seemed to be mutually exclusive. In order to achieve high sensitivities, the conductivity within the microstructure sensor layer must change dramatically when vibrations are detected. To achieve wide sensing range, however, the sensing layers should be constantly conductive even under high tensile strain. High sensitivity and wide range of strain detection seemed impossible for sensors made of one conductive medium.

Bi and the Tsinghua University group devised an approach to simultaneously realize both desired features.

The membrane’s carbon nanofibers are aligned in parallel (pCNF), which gives it a low strain detection limit (and high sensitivity) and a wide range of strain detection. A randomly aligned CNF membrane (rCNF), however, has a wider range. Researchers created a flexible strain sensor that is sensitive and can detect strains in a wide range of strains by stacking parallel and randomly aligned carbon-nanofiber membranes.

Bi stated that the obtained p/r CNF-based strain sensor had a remarkable strain detection limit of 0.005%, and an ultra-high gauge factor value up to 1272 for strains below 0.5%. “Its maximum strain detection limit is 100 percent, which meets the requirements for most human movements.”

The sensor can accurately detect large motions like joint bending as well as minor movements such as facial expression, eye movement, pulse, and speaking.

They created an intelligent system to recognize lip-languages using p/r CNF strain sensors, Arduino, and a loudspeaker as proof of concept. The system “reads lips”, can correctly track phonetic symbols through lip movements and executes the corresponding instructions such as audio or output lights.

Bi said that “The recognition system can help people who have language disabilities, proving its potential in health management.

The current lip-language recognition system cannot handle certain communication situations or locations.

Bi said, “We will create application scenarios for the lip-language recognition software, and improve comfort and portability of wearing.” “We hope such a wearable device could become a second-mouth for people with vocal chord damage and minimize the effects of this injury on someone’s day to day life.

Other high-performance sensors may also benefit from the dual-alignment design of the p/r CNF strain sensor.