Artificial skin based on flexible optical tactile sensors

Artificial skin consists of a surface containing a large number of sensors.¹ To match the human skin as closely as possible, we are developing a range of artificial-skin concepts based on optical pressure sensors that are integrated in flexible or even stretchable foils. This enables sensing of pressure changes on irregular or moving surfaces. For instance, robots equipped with these foils can be given an improved sense of touch. In addition, a large number of medical applications are targeted, such as monitoring pressure distributions in shoe soles or prostheses.

In the last decade, pressure sensors have already been developed to create artificial skin based on electrical sensors on rigid or (in the best case) flexible substrates.^2,3 Optical sensors have a number of advantages compared to their electrical counterparts. They do not suffer from electromagnetic interference and can have higher resolution, speed, and sensitivity. Optical sensors can also be used in harsh environments or places were electrical currents are undesirable, like in the human body.

We are looking into different approaches for artificial-skin development for a range of potential applications. When large sensing areas are needed, we employ fiber-Bragg-grating-based sensors (see Figure 1). To embed fibers into a sensing skin, we use flexible and stretchable host materials (such as silicones). The fiber sensors can be embedded as meanders in a stretchable material to achieve stretchability without damaging the fibers. On the other hand, for applications requiring high-density pressure sensors, we are developing a concept for a novel optical-sensor type based on stacked arrays of crossing waveguides (see Figure 2). Each waveguide crossing is a sensor point. The two layers are separated when the sensor is idle. When pressure is applied on a sensor point, the distance between the waveguides decreases and power is transmitted.

Figure 1. Principle of a sensor based on fiber-Bragg gratings. When we send light covering a broad spectrum into the fiber, light of a particular wavelength (λ_B) is reflected, with the precise wavelength depending on the grating period. Through external influences, such as pressure, this period changes, causing a change in wavelength of the reflected light. λ: Wavelength. P_{opt, in}, P_{opt, refl.}, P_{opt, trans.}: Input, reflected, and transmitted optical power. Λ: Waveguide separation. L: Length.

Figure 2. Principle of our array waveguide sensor. Two layers of polymer waveguides form a grid of crossing points. A force exerted on a crossing point of two waveguides will couple light from an input to an output waveguide.

Fibers in Bragg gratings are widely used as sensors in different commercial applications.⁴ A fiber-Bragg grating acts as a sensor since the wavelength it reflects depends on external parameters such as temperature, strain, and pressure (see Figure 1). Different novel fiber sensors are being designed and fabricated by our partners in the research projects FAOS⁵ and Phosfos.⁶ Our goal is to optimize the responsivity of the fiber-Bragg gratings with respect to inputs such as external pressure or lateral strain. We use microstructured fibers⁷ to reduce the temperature dependence and enhance their sensitivity to pressure. We are also investigating and characterizing polymer optical fibers (POFs), a potentially low-cost alternative. Using POFs as sensing fibers provides a less brittle solution, allowing bending and even stretching of the fibers.

We are investigating different molding techniques to accurately control fiber positioning in the embedding materials⁸ (commercially available silicone or novel polymers developed by our project partners). They enable adjustment of the mechanical parameters and thus ensure biocompatibility (see Figures 3 and 4). We are also working towards integrating optoelectronic components (such as laser or superluminescent diodes) in the embedding materials, since they are needed to drive the fiber sensors. The result is a thin sensing patch that can be attached to or wrapped around an irregularly shaped object such as a robot arm (see Figure 5). Embedding the fibers in a flexible foil also enhances the spatial responsivity of the Bragg grating. Pressure effects can be observed up to a few centimeters from the grating.

Figure 3. Embedded polymer optical fiber in stretchable polydimethylsiloxane material.

Figure 4. Silica single-mode fiber embedded in newly developed flexible materials.

Figure 5. Sensing patch with integrated driving electronics and sensing fiber.

As part of the FAOS project, we are also developing a completely new concept for high-density sensors based on arrays of crossing waveguides. The basic step for fabrication of such sensors is integration of polymer waveguides into silicone material. We realize this by patterning waveguides with lithography on top of a silicone layer (see Figure 6). To separate two layers of waveguides, we use an isolation layer of soft silicone⁹ or manufacture supporting structures between the waveguides. We have fabricated such samples and measurements are now in progress. To test the sensor, we sent infrared light (850nm) into an input waveguide while the output waveguides were observed under different pressure conditions. Figure 7 shows that light is coupled between waveguides when pressure is applied. Further characterization of the sensor is scheduled for the near future. To date, optical sources and detectors have not yet been incorporated, but we are working on the integration of optoelectronics in flexible substrates¹⁰ to obtain a stand-alone flexible sensor foil.

Figure 6. Cross-section of polymer waveguides, patterned lithographically on top of a silicone layer.

Figure 7. CCD view of a cross-section of the array waveguide sensor. (left) Light was sent into the lower waveguides, without applying pressure to the sensor. (right) Pressing the sensor results in light coupling to a waveguide in the upper layer.

In summary, we are developing different types of optical pressure sensors for artificial-skin applications. One type is based on embedded fibers with Bragg gratings for use as large-area sensing foils. The other, based on crossing waveguides, can be used when high-density pressure sensors are required. Both approaches target highly flexible and even stretchable sensor skins. Therefore, we are using embedding materials such as silicones that resemble the human skin.

This work is conducted partially within the framework of the FAOS (funded by the Institute for the Promotion of Innovation by Science and Technology, IWT, Flanders, Belgium) and Phosfos (funded within the European Commision's Seventh Framework Programme). Jeroen Missinne is supported by a PhD fellowship from the Research Foundation of Flanders (FWO-Vlaanderen), while Bram Van Hoe is funded by a grant from IWT-Vlaanderen. The authors also acknowledge support from their colleagues Erwin Bosman, Geert Van Steenberge, Peter Van Daele, and Jan Vanfleteren.