Tactile pressure sensors are used in everything from user interfaces for computers to sensing systems for robots. While they are most readily recognized in trackpads on laptop computers, similar technologies can give robots information about what they are touching, and how much pressure is being applied. But the traditional electronic technologies that make up these devices are not suitable for every use case. In healthcare settings, especially near an MRI machine, strong magnetic fields can render these systems useless. Extreme environments — such as those with high levels of radiation or explosive gasses — are also unsuitable for tactile pressure sensors.
A team of researchers at Tampere University in Finland was inspired by a type of soft actuator that is commonly found in soft robots. These actuators use either fluid or gas to expand chambers in soft materials to trigger movement. They realized that this process could essentially be reversed, and deformations of the actuator would trigger airflow back in the direction of the reservoir. By measuring changes in pressure, one could build a tactile pressure sensor that is soft and completely devoid of electronic components.
The design of the new sensor (📷: V. Lampinen et al.)
A single pressure sensor can only provide just so much information, so the team developed a soft touchpad sensor that consists of 32 pneumatic channels arranged in a grid pattern with 16 channels aligned row-wise and 16 column-wise. When a force is applied to the pad, only the row and column channels nearest to the contact location deform, leading to an increase in their flow resistance, especially in the channel directly beneath the touch point.
The touchpad measures this flow resistance through a pressure divider system, analogous to an electrical voltage divider, where each pneumatic channel connects in series with a constant pneumatic resistor. A constant supply pressure of 60 kPa is applied on one side, while the opposite side vents to atmospheric pressure. By measuring the pressure differences at specific junctions along each channel using pressure gauges, the touchpad accurately detects the touch location and intensity.
Fabricated from soft polydimethylsiloxane using microfabricated molds, each channel has a 200 micrometer by 200 micrometer cross-section and spans 40 mm in length, with 2.65 mm spacing between channels, resulting in a 40 mm x 40 mm active sensing area within a 68 mm x 68 mm device.
Pressure maps produced by the new sensor (📷: V. Lampinen et al.)
A study demonstrated that the sensors maintain their functionality even when bent, with only slight variations in pressure response due to channel deformation. However, ambient conditions such as air pressure, temperature, and humidity do affect the sensor’s performance. For instance, lower supply pressures reduce sensitivity, but this effect can be normalized. Similarly, increased temperatures slightly decrease sensitivity due to changes in air properties and elastomer characteristics, while higher humidity also reduces sensor response. Despite these influences, the sensor’s performance was stable in typical laboratory conditions.
However, the touchpad still has limitations when compared to modern capacitive touchpads, which use row-column scanning to independently sense each crossing, allowing for much higher resolutions and the detection of numerous simultaneous touches. Unlike capacitive touchpads that can scan a large matrix at frequencies of 200 Hz or more, fluidic systems face challenges in achieving such high scanning rates because pneumatic valves cannot switch rapidly enough. As it stands, this technology is not the best in class. But in some cases, it may be the only viable option, and it has been shown to perform reasonably well for many applications.
